SPECTROSCOPY AND SPECTRAL ANALYSIS

News

Forgot your password?

Enter your e-mail address to receive an e-mail with your account information.

Journal Online

Archive

Links

	CAST

	CNCOS

	CNKI

	WanfangDATA

	CHEMSOC

	sinospectroscopy

	CPS

Quick Search

Adv Search

2023 Vol. 43, No. 01 Published: 2023-01-01

	1	The Mechanism of Hydrocarbon Flame Soot Formation in Spectral Diagnosis: A Review
		WU Run-min¹, XIE Fei¹, SONG Xu-dong^1*, BAI Yong-hui¹, WANG Jiao-fei¹, SU Wei-guang¹, YU Guang-suo^{1, 2}
		DOI: 10.3964/j.issn.1000-0593(2023)01-0001-08
		As the main by-product of incomplete combustion of hydrocarbon fuel, soot not only has a detrimental effect on the combustion equipment but also harms air quality and human health. The formation of soot is a complex physical and chemical process. It is necessary to decrease the striking differences between the physical evolution and the chemical process of soot generation and combustion. In the past few decades, some scholars have investigated the mechanism of soot generation by spectral diagnostic technology and have had the understanding of soot formation in different generation stages to some extent. By using spectral diagnosis technology, the process of soot generation in the combustion system could be researched more comprehensively, and the molecular composition, fine structure, concentration distribution and other characteristics of soot could also be determined. More importantly, the process could be explained from the changes of soot structure and blackbody radiation intensity in detail. The aim of this study, on the one hand, is to illustrate the research process and developing trend of hydrocarbon flame soot by using spectral diagnosis technology. On the other hand, it is to discuss the problems like the accuracy of the radiation intensity generated during the process of soot generation in flames containing background radiation by the diagnostic tools, LIBS, LII and LIF included. Additionally, the formation mechanism of hydrocarbon flame soot (from precursor generation and growth to particle generation and aggregation, and finally, particle oxidation) is introduced in detail. Meanwhile, it summarized the application of spectral diagnostic methods to detect the properties of soot and the research status of spectral diagnostic techniques on the characterization of soot during combustion, including the volume fraction, reaction temperature and structure characterization of soot based on image processing, and the influence of polycyclic aromatic hydrocarbons, reaction atmosphere and temperature on the formation of soot particles was explored as well. Finally, the future application of spectral diagnostic methods in soot is forecasted. Spectroscopy diagnosis will present a more detailed and accurate study of the chemical reaction mechanism of soot formation in the future. The impact of uneven soot on flame images needs to be reduced, and the development trend of optimizing spectral diagnostic measurement methods to collect and monitor the concentration of multiple gas components in flame simultaneously and the generated soot in real-time. The analysis of spectral diagnostic and image analysis in homogeneous combustion flame soot will provide directions and have crucial scientific significance for promoting clean combustion and heterogeneous flow research.
		2023 Vol. 43 (01): 1-8 [Abstract] ( 148 ) RICH HTML PDF (4879 KB) ( 212 )

	9	Advances and Prospects in Inner Surface Defect Detection Based on Cite Space
		SHENG Qiang^{1, 2}, ZHENG Jian-ming^1*, LIU Jiang-shan², SHI Wei-chao¹, LI Hai-tao²
		DOI: 10.3964/j.issn.1000-0593(2023)01-0009-07
		In order to analyze the development, trend and dynamics of inner surface defect detection, 4 708 relevant literature in English and 818 in Chinese were collected through the search of relevant literature in this field in WoS and CNKI databases. The visual analysis software CiteSpace is used to study the knowledge map of literature co-occurrence and clustering, analyze the distribution status and cooperation of internal surface defect detection in countries, institutions and scholars, and sort out the research hotspots and cutting-edge trends. It is found that the research on inner surface defect detection has obvious interdisciplinary attributes, mainly involving analytical chemistry, material science, spectroscopy, instrumentation, mechanical engineering and computer science. In recent years, the annual growth rate of related literature in the WoS database has been more than 10%, and the annual growth rate of CNKI has been more than 20%. China and the United States have become the most active countries in this field, accounting for about 40% of the total number of publications. Chinese scholars’ research in non-destructive testing, image processing and other fields lags behind that of foreign scholars, but they catch up in machine vision and deep learning. According to the research route, it can be divided into detection based on acousto-optic electrothermal magnetism and detection based on the visual imaging. The former includes the acquisition of spectral, ultrasonic and electromagnetic images by different technical means and the realization of defect detection by image processing technology, while the latter is the main defects recognition and classification based on visual image, has become the main research focus in the field. The development of inner surface defect detection can be divided into three stages: defect identification, defect classification and defect analysis. Before 2000 defects were recognized and determined mainly by thermal, acoustic, optic, electrothermal, and magnetic signals or images. Since 2000, the support vector machine (SVM) technology greatly improves the efficiency and accuracy of defect classification. In recent ten years, with the increasing demand for defect analysis and measurement, defect location and measurement based on machine vision has gradually become a development trend, and the object of defect detection has gradually developed to the inner surface of deep holes and small holes.
		2023 Vol. 43 (01): 9-15 [Abstract] ( 125 ) RICH HTML PDF (2471 KB) ( 194 )

	16	Progress of the Application of MIR and NIR Spectroscopies in Quality Testing of Minor Coarse Cereals
		FENG Hai-zhi¹, LI Long^1*, WANG Dong², ZHANG Kai¹, FENG Miao¹, SONG Hai-jiang¹, LI Rong¹, HAN Ping²
		DOI: 10.3964/j.issn.1000-0593(2023)01-0016-09
		Mid-infrared spectroscopy and near-infrared spectroscopy are important analysis methods in modern analytical chemistry, which are the important technical means for human beings to recognize the structure, function, components and content of substances. Minor coarse cereals generally refer to a variety of grains and beans with a short growth period, small planting areas, special planting areas and methods, and special purposes, which are characterized by small, small, special and miscellaneous. Being rich in nutrition, minor coarse cereals are not only traditional rations but also healthy food resources. With the improvement of people’s living standards and dietary structure, minor coarse cereals, as a new food resource with the same origin as medicine and food, play an important role in modern green health food. The quality detection of minor coarse cereals will provide reliable data support for the research of bioactive substances, quality classification, minor coarse cereals breeding, origin traceability and authenticity identification of small cereals. The literature on quality testing of minor coarse cereals in China in recent 30 years were summarized according to the classification of wheat minor coarse cereals and legume minor coarse cereals. It was demonstrated by the research result that there are more literature on the quality testing of wheat minor coarse cereals, accounting for about 2/3 of the literature, and the application of near-infrared spectroscopy is the majority; there are relatively little literature on the quality testing of legume minor coarse cereals, accounting for about 1/3 of the literature, and the application of mid-infrared spectroscopy is the majority. Many important applications in the quality testing and analysis of minor coarse cereals were carried out by mid-infrared and near-infrared spectroscopy. Among them, mid-infrared spectroscopy is more applied to characterize the active substances in minor coarse cereals and the processing of minor coarse cereals. However, near-infrared spectroscopy is more applied to the quantitative analysis and testing of main qualities such as crude protein, crude fat and moisture in minor coarse cereals, which can provide an efficient data source for minor coarse cereals quality monitoring and scientific breeding. In recent years, with the development of chemometrics and the progress of computer technology, near-infrared spectroscopy is no longer limited to the quantitative analysis of qualities of minor coarse cereals but is also applied to the traceability of it, which has achieved good results. Finally, the application of mid-infrared and near-infrared spectroscopy in non-destructive analysis and testing of small grain quality has prospected.
		2023 Vol. 43 (01): 16-24 [Abstract] ( 149 ) RICH HTML PDF (877 KB) ( 315 )

	25	Identification of the Origin of Bluish White Nephrite Based on Laser-Induced Breakdown Spectroscopy and Artificial Neural Network Model
		BAO Pei-jin¹, CHEN Quan-li^{1, 3*}, ZHAO An-di¹, REN Yue-nan²
		DOI: 10.3964/j.issn.1000-0593(2023)01-0025-06
		To promote the application of artificial neural network technology in identifying the origin of gems, an artificial neural network model of semi-quantitative trace element content of bluish white nephrites obtained by laser-induced breakdown spectrometer was established. The element content data were obtained by laser-induced breakdown spectrometer in the uniform and clean parts of nephrites from six regions: Xinjiang, Guangxi, Jiangsu, Qinghai, Korea and Russia. After screening using data filtering principles and normalizing the data, the collinearity between data is discussed by factor analysis and linear regression, and the discriminant model of the artificial neural network is established. The results show that the VIF value of each selected variable is less than 5, so there is no obvious multicollinearity among the selected elements. The KMO value of factor analysis is less than 0.6, indicating that there is no obvious relationship between variables. Moreover, thet-SNE graph of nephrite is used to reduce and visualize the data. T-SEN graph shows that most of the data points are overlapped together, indicating that the data’s simple clustering and correlation analysis could not distinguish the origin. Therefore, the artificial neural network is selected for the identification analysis of the six origin data. After the iterative discrimination of the artificial neural network model, the accuracy of the model for the identification of the blue and white nephrite from six producing areas is up to 0.933, and the nephrite from Korea has the highest data discrimination accuracy of 0.995 with an error of 0.028,while nephrite from Qinghai has the lowest data discrimination accuracy of 0.803 with an error of 0.090. In conclusion, a laser-induced breakdown spectrometer combined with the artificial neural network has great potential in applying gem origin tracing.
		2023 Vol. 43 (01): 25-30 [Abstract] ( 82 ) RICH HTML PDF (2879 KB) ( 208 )

	31	Research on the Spectral Prediction Model of Gravure Spot Color Scale Based on Density
		HAI Jing-pu^{1, 2}, GUO Ling-hua^{1, 2*}, QI Yu-ying^{1, 2}, LIU Guo-dong^{1, 2}
		DOI: 10.3964/j.issn.1000-0593(2023)01-0031-06
		A model for predicting spectral reflectance of gravure spot color scales is proposed based on the relationship between density and spectral reflectance. Firstly, this method establishes the relationship between the solid spectral reflectance and its density according to the definition of density, and establishes the calculation method of the tone spectral reflectance based on the solid spectral reflectance. Then, according to the superposition principle of density, assuming that the proportional relationship between tone density and solid density is established, the relationship between tone density solid density and substrate density is established. Finally, combined with calculating the tone spectral reflectance, a spectral reflectance prediction model of the gravure spot color scale is established. Thirty kinds of spot color inks are mixed and used to print samples by gravure printing. The prediction model is verified by the coefficient of determination R² and the color difference. The experimental results show that the proportional coefficients of the actual tone density and the actual solid density are the same under the same dot area rate of different spot colors, and the determination coefficients R² of both are greater than 0.98. Based on this relationship, the prediction model established in this paper has a high determination coefficient at different dot area rates, the root mean square error is less than 0.01, and the maximum color difference is 2.667 NBS. Finally, another ten different spot color inks are prepared to print simples under the same process conditions, and the proportion coefficient between the actual tone density and the actual solid density is used to verify the accuracy of the model in predicting the tone spectral reflectance of spot color ink by color difference formula. The color difference results show that 82.12% of the color difference is less than 2.5 NBS, most of the color difference is between 0.5～2 NBS, accounting for 58.32% of the total frequency, and the average color difference is 1.58 NBS, which meets the requirements of enterprises for fine reproduction of colors. It is verified that the model has high accuracy for predicting the spectral reflectance of the gravure spot color scale, and it can provide a scientific method for digital proofing instead of gravure machine proofing.
		2023 Vol. 43 (01): 31-36 [Abstract] ( 88 ) RICH HTML PDF (2543 KB) ( 73 )

	37	Research on the Factors Influencing the Non-Destructive Detection of Potatoes by Near-Infrared Spectroscopy
		HAN Min-jie, WANG Xiang-you, XU Ying-chao^*, CUI Ying-jun, LÜ Dan-yang
		DOI: 10.3964/j.issn.1000-0593(2023)01-0037-06
		To enhance the stability as well as accuracy of near-infrared rapid nondestructive testing, this paper compares the spectral effects of three different types of light sources in the conditions of fiber optic light source, halogen cup light source and ring light source, and the results show that the ring light source has the lowest spectral noise, moderate irradiation intensity and the best uniformity. Based on these studies, the power of the light source, the distance from the optical source to the potato surface, and the distance from the optical fiber to the detection point on the potato surface are investigated in this study. The spectral model’s prediction effect on potatoes’ soluble solids content under different factors was evaluated by a three-factor, three-level response surface test. The optimal parameters were light source power 238.33 W, distance from fiber optic probe to sample surface 8.17 mm, distance from the light source to sample surface 370 mm, R_P=0.867, and RMSE_P=0.149°Brix of the PLSR model for soluble solids prediction. To further eliminate equipment and environmental noise, interference of noise was reduced by different preprocessing algorithms. The outcome showed that the algorithm with the standard variable ranking approach had the best noise removal effect, with R_P=0.914 and RMSE_P=0.132°Brix, which obtained a better prediction result while effectively removing noise. The test results show that optimizing the testing environment and conditions through response surface testing can effectively improve the prediction accuracy of potato quality testing and provide technical guidance for the construction and device selection of near-infrared online the nondestructive detection environment for potatoes.
		2023 Vol. 43 (01): 37-42 [Abstract] ( 114 ) RICH HTML PDF (3197 KB) ( 126 )

	43	Research on Long Optical Path and Resonant Carbon Dioxide Gas Photoacoustic Sensor
		LI Zhen-gang^{1, 2}, SI Gan-shang^{1, 2}, NING Zhi-qiang^{1, 2}, LIU Jia-xiang¹, FANG Yong-hua^{1, 2*}, CHENG Zhen^{1, 2}, SI Bei-bei^{1, 2}, YANG Chang-ping^{1, 2}
		DOI: 10.3964/j.issn.1000-0593(2023)01-0043-07
		Carbon dioxide (CO₂) is the raw material of plant photosynthesis and greenhouse gas. Its excessive emission will affect the ecological environment of animals and plants. Under the background of carbon peaking and carbon neutrality, it is of great significance to develop high-sensitivity CO₂ detection devices. In order to monitor the change of CO₂ concentration in the atmospheric environment, a long optical path and resonant CO₂ gas photoacoustic sensor was designed, and a photoacoustic detection setup was built. Adistributed feedback laser (DFB) with a central wavelength of 2 004 nm was used as the excitation light source. The laser entered into a spherical absorption cell made of diffuse reflective material, and multiple reflections occurred in the cell to increase the absorption path of the gas. To reduce the thermal noise generated by the absorption of light energy by the absorption cell, the outside of the cell was wrapped by two aluminum hemispheres with high thermal conductivity. An acoustic tube was coupled with the absorption cell. When the tube worked in the first-order longitudinal resonance mode, the photoacoustic signal was amplified and reached the maximum at the end of the tube. The CO₂ relaxation rate was greatly accelerated, and the thermal, acoustic conversion efficiency was improved by saturated humidifying the sample, which further amplified the photoacoustic signal. The photoacoustic signal produced by the humidified sample was about 2.1 times that of the dry sample. The photoacoustic detection setup was calibrated with a series of wet CO₂ samples, and the results showed a good linear relationship between photoacoustic signals and concentrations. On this basis, the accuracy and stability of the setup were verified through the detection experiment of standard gas. Allan variance was used to evaluate the detection sensitivity of the setup under long-time operation. When the average time was 865 s, the detection sensitivity was ~0.35×10^-6. Compared with the traditional T-type photoacoustic cell, the optical path was increased by~20 times, and the photoacoustic signal was amplified by ~6 times. The setup was used to detect CO₂ in the outdoor environment for 10 hours, and the average concentration of outdoor CO₂ was ~381×10^-6. In conclusion, due to the combination of a long optical path, acoustic resonance and humidified samples, the photoacoustic signal of CO₂ was effectively increased, which provided a relevant reference for the design of gas photoacoustic sensor and detection setup.
		2023 Vol. 43 (01): 43-49 [Abstract] ( 96 ) RICH HTML PDF (4441 KB) ( 134 )

	50	Quantitative Analysis of Hemoglobin Based on SiPLS-SPA Wavelength Optimization
		GAO Xi-ya^{1, 2, 3}, ZHANG Zhu-shan-ying^{1, 2, 3*}, LU Cui-cui^{1, 2, 3}, MENG Yong-ji^{1, 2, 3}, CAO Hui-min^{1, 2, 3}, ZHENG Dong-yun^{1, 2, 3}, ZHANG Li^{1, 2, 3}, XIE Qin-lan^{1, 2, 3}
		DOI: 10.3964/j.issn.1000-0593(2023)01-0050-07
		Hemoglobin is an important physiological index of the human body. Abnormal concentrations of Hemoglobin will lead to various diseases. Infrared spectroscopy has the advantages of simplicity, non-destructive and rapidity. It is very suitable for the quantitative analysis of physiological parameters. However, the spectral background is complex, and the effective information is weak. How extract effective feature variables and build an accurate quantitative model is a difficult problem. To solve this problem, study the spectral data of blood samples and hemoglobin imitation solution samples, and through modeling and comparison, the best data set division method is SPXY by using SPXY method, K_S method, duplex method and equal interval division method to divide the data. Three data pre-processing methods of Savitzky Golay first-order derivative filter (S_G1) + wavelet transform, wavelet transform+S _G1 and standard normal variable transform (SNV)+S_G1 are traversed, and the best pre-processing method is SNV+S_G1. Combined with the series idea, the characteristic wavelength optimization method of combining Synery interval Patial Least Squares (SiPLS) and Successive Projections Algorithm (SPA) in series is proposed, and so the SiPLS-SPA-PLS prediction model is constructed. The model is verified with two data sets, and the advantages and disadvantages are judged according to the evaluation indexes and compared to the three quantitative models of full spectrum PLS, SPA-PLS and SiPLS. The experimental results show that: (1) using SiPLS-SPA-PLS for quantitative analysis, the values of R_C, R_P, RMSEC and RMSEP of blood samples are 0.993 6, 0.990 6, 0.199 2 and 0.184 6 respectively, and the values of R_C, R_P, RMSEC and RMSEP of imitation solution samples are 0.998 9, 0.998 5, 1.848 9 and 2.007 4 respectively. Compared with the three quantitative models of full spectrum PLS, SPA-PLS and SiPLS, the SiPLS-SPA-PLS model is the best. Because the values of R_C and R_P are the largest and the values of RMSEC and RMSEP are the smallest. This model can realize the quantitative analysis of hemoglobin more accurately. (2) The SiPLS-SPA-PLS quantitative model can screen the optimal wave band more accurately. The effective wave bands screened by the two samples are blood (1 144~1 264, 1 606~1 798 nm) and imitation solution (1 018~1 390, 1 600~1 700 nm). The influencing factors of the instrument are roughly the same. This method can accurately optimize the characteristic wavelength. (3) The model can extract effective variables, remove the influence of useless noise, select 28 spectral variables from 700 blood samples and 41 spectral variables from 1 201 hemoglobin imitation solution samples to improve the detection speed and prediction efficiency. In short, this method provides an idea for rapid and accurate detection of hemoglobin.
		2023 Vol. 43 (01): 50-56 [Abstract] ( 94 ) RICH HTML PDF (2883 KB) ( 204 )

	57	Measurement and Analysis of Uranium Using Laser-Induced Breakdown Spectroscopy
		ZHANG Zhi-wei^{1, 2}, QIU Rong^{1, 2*}, YAO Yin-xu^{1, 2}, WAN Qing³, PAN Gao-wei¹, SHI Jin-fang¹
		DOI: 10.3964/j.issn.1000-0593(2023)01-0057-05
		To promote the application of LIBS technology in the detection of trace heavy metals and nuclear pollution detection, and improve the sensitivity and accuracy of detection, laser double-pulse LIBS technology and photoelectric double-pulse LIBS technology were used to analyze the uranium elements in soil and silica respectively. Firstly, optimizing the laser pulse energy, voltage and acquisition delay parameters to improve the intensity and signal-to-noise ratio of the characteristic spectrum of uranium; Then, under the conditions of optimized experimental parameters, the soil samples and silica samples containing different concentrations of uranium were analyzed. Two characteristic spectral lines of uranium element UII 367.01 nm and UII 454.36 nm were selected as the analysis line, and the calibration curve was established through the linear relationship between the uranium element concentration and the characteristic spectral line intensity. Under the condition of laser double pulse excitation, laser 1 was used as the pre-pulse, the main parameters were 1 064 nm, 90 mJ, 9.2 ns, and laser 2 was used as the reheating pulse, the main parameters are 355 nm, 50 mJ, 8 ns. The time interval between the two pulses is 800 ns, and the spectral acquisition is delayed by 1 μs, the lower detection limits of the concentration of uranium in the soil and silica samples were 572 and 110 mg·kg^-1, respectively, and the goodness of fit value R² were respectively 0.958 and 0.999. Under the condition of photoelectric double pulse excitation, the laser pulse was used as the pre-pulse, the main parameters were 355 nm, 50 mJ, 8 ns, the high-voltage electric pulse was used as the reheating pulse, the main parameters were 3 900 V, square wave, pulse width 50 μs, the time interval of the two pulses 1 μs. The lower detection limits of the concentration of uranium in soil and silica samples were 108 and 64 mg·kg^-1, and the goodness-of-fit values R² were 0.991 and 0.997, respectively. The research results show that under the same excitation conditions, the characteristic spectrum of uranium has an obvious matrix effect, which has higher spectral intensity, lower detection limit and higher goodness of fit value in silica samples. Compared with double laser pulse, the photoelectric double pulse can significantly enhance the intensity, stability and signal-to-noise ratio of the characteristic spectrum of uranium element, and the optical path of the photoelectric double pulse system is simpler has an important reference for the development and application of LIBS technology significance. The research method and results can provide technical support for detecting heavy metal pollution in soil and the emergency monitoring of soil and aerosol in the event of nuclear leakage.
		2023 Vol. 43 (01): 57-61 [Abstract] ( 138 ) RICH HTML PDF (2037 KB) ( 253 )

	62	Study on Identification Seawater Submersible Oil Based on Total Synchronous Fluorescence Spectroscopy Combined With High-Order Tensor Feature Extraction Algorithm
		KONG De-ming¹, CUI Yao-yao^{2, 3}, ZHONG Mei-yu², MA Qin-yong², KONG Ling-fu²
		DOI: 10.3964/j.issn.1000-0593(2023)01-0062-08
		Submersible oil is a kind of oil spill hidden under the sea surface in a suspended state. It has poisoned and eroded the marine ecological environment for a long time. However, effective monitoring means and treatment methods have not been formed for submersible oil pollution, which makes its pollution more sudden and harmful than a sea oil spill. Therefore, it is of great significance to studying effective submersible oil identification methods to protect the marine ecological environment. The TSFS in three-dimensional fluorescence spectroscopy has the advantages of no Rayleigh scattering interference and less redundant data in detecting and identifying oil pollutants. The application of the multidimensional correction analysis method to TSFS data is limited because it does not have a trilinear structure. Thus, a new identification method for seawater submersible oil samples was proposed by combining TSFS with a high-order tensor feature extraction algorithm. First, 90 submersible samples were prepared by using organic dispersants and six different kinds of oil products. Then, the TSFS data of samples were collected using an FS920 fluorescence spectrometer, and the data were preprocessed by standardized. Finally, the identification models of submersible oil samples were established by 2D-LDA and 2D-PCA in the high-order tensor feature extraction method. The established model was compared with the identification model established by conventional MCR-ALS-LDA and NPLS-DA. The results show that the submersible oil sample identification models established by 2D-LDA and 2D-PCA have robust and reliable performance, and the accuracy, sensitivity and specificity of the identification models were 100%, 100% and 100%, respectively. In addition, the fine spectral features of the TSFS spectral image matrix in space, statistics, and graphics can be directly extracted by 2D-LDA and 2D-PCA, which brings a more accurate identification basis for distinguishing submersible oil samples. Therefore, compared with the conventional methods based on expansion or decomposition of data, the more accurate prediction results were obtained by the discrimination model established by the high-order tensor feature extraction method. This study provides a reference for submersible oil identification.
		2023 Vol. 43 (01): 62-69 [Abstract] ( 83 ) RICH HTML PDF (5079 KB) ( 53 )

	70	A Study on the Thermal Infrared Spectroscopy Characteristics of the Skarn Minerals in Zhuxi Tungsten Deposit, Jiangxi Province
		FU Ming-hai^{1, 2}, DAI Jing-jing^1*, WANG Xian-guang³, HU Zheng-hua⁴, PENG Bo¹, WAN Xin³, ZHANG Zhong-xue², ZHAO Long-xian^{1, 2}
		DOI: 10.3964/j.issn.1000-0593(2023)01-0070-08
		The mineralization of scheelite in the Zhuxi tungsten deposits is closely related to skarnization. Scheelite is mostly produced together with skarn minerals such as garnet and diopside. In this study, for the first time, the typical skarn minerals such as garnet, diopside, vesuvianite, wollastonite and actinolite are measured by micro infrared spectroscopy and electron probe analysis to explore the thermal infrared spectral characteristics of skarn minerals in Zhuxi and their implications for mineralization, and to establish a thermal infrared spectral library of skarn minerals in Zhuxi area. The results show that the garnet in Zhuxi tungsten deposit is mainly grossular-and radite series. There are two absorption peaks (a large and a small) near 800 and 920 cm^-1, and there is a characteristic absorption valley near 880 cm^-1. When the grossular content in garnet is greater than 50%, the characteristic absorption valley of garnet is located at 880~900 cm^-1. When grossular content is less than 50%, the absorption valley of garnet is located at 865~875 cm^-1. With the increase of Al₂O₃ content in garnet, the characteristic absorption valley moves towards the high-wavenumber direction. Grossular tends to high wavenumber while andradite tends to low wavenumber; Diopside is mainly diopside-hedenbergite series. There is a diagnostic step-shaped absorption peak in the wavenumber range of 850~950 cm^-1, anabsorption peak at 1 050 cm^-1 and a weak double-valley absorption at 1 000 cm^-1. With the decrease of diopside content, MgO content decreases the diopside absorption peak moves to the low-wavenumber direction. The absorption peak of hedenbergite is concentrated in the low-wavenumber range compared with diopside, which is consistent with the changing law of garnet. It is speculated that the reason is that Al and Mg are more active than Fe. Vesuvianite has similar absorption peaks in the range of 850~950 cm^-1 as diopside. The difference is that vesuvianite still has anabsorption peak at 800 cm^-1; Wollastonite has three absorption peaks (a large and two small) near 875, 1 000 and 1 060 cm^-1, and two characteristic absorption valleys near 980 and 1 040 cm^-1; Actinolite has two absorption peaks (a small and a large) near 750 and 900 cm^-1, and three characteristic absorption valleys near 770, 930 and 1 020 cm^-1.The scheelite mineralization in the Zhuxi deposit is most closely related to garnet and diopside, and it mainly grows in veins along the boundary of garnet and diopside. Their thermal infrared spectra can be used as an indicator for searching for scheelite. The above results have theoretical and practical significance for in-depth analysis and research on the mineralogy characteristics and genetic environment of the Zhuxi tungsten deposit in Jiangxi, as well as for exploring the possibility of using thermal infrared technology to guide skarn mineral zoning and ore prospecting.
		2023 Vol. 43 (01): 70-77 [Abstract] ( 98 ) RICH HTML PDF (3595 KB) ( 91 )

	78	Quantitative Analysis of Single Component Oils in Quinary Blend Oil by Near-Infrared Spectroscopy Combined With Chemometrics
		HU Xiao-yun¹, BIAN Xi-hui^{1, 2, 3*}, XIANG Yang², ZHANG Huan¹, WEI Jun-fu¹
		DOI: 10.3964/j.issn.1000-0593(2023)01-0078-07
		The rapid and accurate quantitative analysis of blend oil is of great importance for the quality control of blend oil. However, most previous studies on the quantitative analysis of blend oil have focused on binary, ternary and quaternary blends, and few studies have been conducted on more multi-component blend oil, which is difficult to meet the needs of blend oil detection. This study explores the feasibility of near infrared spectroscopy combined with chemometrics for the quantitative analysis of the singlecomponentoil in quinary blend oil. 51 quinary blend oil samples were formulated from corn oil, soybean oil, rice oil, sunflower oil and sesame oil, and their NIR spectra were measured in a transmittance mode in the range of 12 000～4 000 cm^-1. Firstly, the sample set partitioning based on joint x-y distances (SPXY) algorithm was used to divide the sample into 38 calibration and 13 prediction set samples. Secondly, the modeling effect of five multivariate calibration methods, including principal component regression (PCR), partial least squares (PLS), support vector regression (SVR), artificial neural network (ANN), and extreme learning machine (ELM), were examined for the quantitative analysis of each component in quinary blend oil. Then six spectral preprocessing methods including Savitzky Golag smoothing(SG smoothing), standard normal variate (SNV), multiplicative scatter correction (MSC), first derivative (1^st Der), second derivative (2^nd Der), and continuous wavelet transform (CWT) were compared based on the best modeling method and the reasons for the effectiveness of the preprocessing methods were discussed. Finally, based on the optimal preprocessing method, the competitive adaptive reweighted sampling (CARS) and Monte Carlo uninformative variable elimination (MCUVE) algorithms were further used to screen the variables associated with the predicted components. The results showed that PLS was the optimal modeling method among the five modeling methods, with root mean square error of the prediction set (RMSEP) of 5.564 4, 5.559 2, 3.592 6, 7.421 8, and 4.193 0 for the five components of corn oil, soybean oil, rice oil, sunflower oil, and sesame oil, respectively. After preprocessing-variable selection and then PLS modeling, the RMSEP for the five components were 1.955 3, 0.562 4, 1.145 0, 1.619 0 and 1.067 1, respectively and the correlation coefficients of prediction set (R_p) were all higher than 0.98, indicating that with appropriate spectral preprocessing, variable selection and modeling methods, the accuracy of quantitative analysis of each component in quinary blend oil was greatly improved. This research provided a reference for rapid and non-destructive quantitative detection of multi-component blend oil.
		2023 Vol. 43 (01): 78-84 [Abstract] ( 161 ) RICH HTML PDF (3171 KB) ( 134 )

	85	Surface Enhanced Raman Spectroscopy Analysis of Fentanyl in Urine Based on Voigt Line
		HE Yao^{1, 2}, LI Wei^{1, 2}, DONG Rong-lu², QI Qiu-jing³, LI Ping⁵, LIN Dong-yue^2, MENG Fan-li⁴, YANG Liang-bao^2
		DOI: 10.3964/j.issn.1000-0593(2023)01-0085-08
		Fentanyl substance abuse and death cases frequently occur worldwide, and the detection and identification of fentanyl substances in the human body are becoming increasingly important. After some time in the human body, some of the fentanyl substances are still discharged with urine, so the history of drug abuse can be reflected by detecting fentanyl substances in the urine. Surface Enhanced Raman Spectroscopy ( SERS ) is fast, sensitive and easy to operate, which is suitable for the field detection and analysis of fentanyl in urine. However, the background peaks of urea and other substances in urine are highly coincident with the SERS characteristic peaks of fentanyl, and the characteristic peaks of fentanyl are covered by the background peaks of urine, which causes great interference with the spectral identification of fentanyl in urine.In this paper, the spectral peak analysis model is established based on the Voigt line, and the spectral peak analysis of the overlapping part of urine and fentanyl is carried out. Because of the interference of SERS spectral noise and fluorescence on the peak analytical model, the unconstrained Nelder-Mead algorithm is used to optimize and calculate the model. The algorithm is insensitive to the initial value of iterative parameters to improve the accuracy of the peak analytical model. According to the characteristics of the half-peak width of the SERS spectrum, the analytical peak set was screened, and the urine background peak of the SERS spectrum was deducted to restore the spectral peak characteristics of the fentanyl SERS spectrum at 1 000 and 1 030 cm^-1.The experimental results and phenomena show that the spectral peak analysis model established by the Voigt line shape has a fitting degree of more than 99% for the SERS spectrum of fentanyl in urine and can restore the SERS spectrum characteristics of fentanyl in urine through the screening of the peak solution set. The characteristics of the half-peak width and peak ratio of the reduced spectrum and the characteristic peaks of fentanyl are highly consistent. When the SERS spectrum of blank urine is analyzed, the analytical peak set does not contain the characteristic peaks of fentanyl substances, which can effectively distinguish blank urine from urine containing fentanyl substances. The reduced spectral fragment (935~1 100 cm^-1) was identified by the hit quality index (HQI), which can effectively distinguish Ofentanyl, Furanyl and Acetylfentanyl in urine and improve the discrimination between spectra. This analytical model is expected to provide a way to solve practical problems for the identification and judgment of fentanyl in urine.
		2023 Vol. 43 (01): 85-92 [Abstract] ( 108 ) RICH HTML PDF (5131 KB) ( 47 )

	93	Inversion Model of Clorophyll Content in Rice Based on a Bonic Optimization Algorithm
		LI Xiao-kai, YU Hai-ye, YU Yue, WANG Hong-jian, ZHANG Lei, ZHANG Xin, SUI Yuan-yuan^*
		DOI: 10.3964/j.issn.1000-0593(2023)01-0093-07
		The accurate, efficient and nondestructive detection of chlorophyll content in rice leaves using spectral information is of practical importance for diagnosing and optimizing nitrogen nutrition in rice leaves, developing and optimizing nitrogen fertilization systems in rice fields, and monitoring and evaluating rice pests and diseases. This paper addresses the problem of poor model accuracy and stability when machine learning models are used solely to invert the chlorophyll content of rice leaves. Moreover, takes Northeast japonica rice Jijing88 as the research object, obtains leaf phenotypic hyperspectral data and relative chlorophyll content at key fertility stages such as tillering through grid tests, select the kernel limit learning machine (Kernel function extreme learning machine, KELM) in machine learning as the base modeling model, and proposes a new idea of selecting preprocessing methods based on the base KELM modeling effect first, and then optimizing the KELM training process corresponding to the selected preprocessing combination using a bionic optimization algorithm to improve the model prediction accuracy. First, this paper investigates the preprocessing methods of spectral data, and a total of 72 preprocessing combinations are obtained by combining all four types of preprocessing methods. The sequential projection algorithm (Successive Projections Algorithm, SPA) is used to select the characteristic bands for input into the KELM model to filter the better preprocessing combinations. Based on the modeling effect, the test set’s coefficient of determination (R²_p) corresponding to KELM for the pretreatment combinations CWT+MMS, CWT+MSC+SG+SS, and CWT+SS was higher, 0.850, 0.835, and 0.828, respectively. Secondly, to make the KELM model perform optimally while ensuring stability and generalization. In this paper, the Harris Hawk Optimization Algorithm (Harris Hawks Optimizer, HHO) is introduced to automatically and optimally adjust the parameters of the above three KELM models by simulating the cooperative behavior and chasing strategy of the hawks during predation, resulting in the HHO-KELM models with R²_p of 0.957, 0.867 and 0.858, respectively, and a maximum of 10.7% effectively improves the model accuracy. The feasibility of the HHO algorithm to optimize the machine learning model to invert the chlorophyll content of rice leaves was demonstrated, which provides a strong reference and reference for the determination and assessment of chlorophyll content in northeastern japonica rice.
		2023 Vol. 43 (01): 93-99 [Abstract] ( 110 ) RICH HTML PDF (2681 KB) ( 235 )

	100	Research on Online Detection of Tea Stalks and Insect Foreign Bodies by Near-Infrared Spectroscopy and Fluorescence Image Combined With Electromagnetic Vibration Feeding
		SUN Xu-dong^{1, 2}, LIAO Qi-cheng¹, HAN Xi³, Hasan Aydin⁴, XIE Dong-fu¹, GONG Zhi-yuan¹, FU Wei¹, WANG Xin-peng¹
		DOI: 10.3964/j.issn.1000-0593(2023)01-0100-07
		Tea is one of the health drinks favored by the consumer, but during the process of tea machine harvesting and processing, it is easy to be mixed with tea stalks and foreign insect bodies. It resulted in pollution and influenced the quality and safety of tea products. In the future, we should focus on preventing and detecting of foreign bodies. X-ray imaging technology, based on the density difference between food substrate and foreign bodies, is widely applied to detect metal foreign bodies and extended to high-density plastic. However, it is not suitable for low-density organic foreign bodies such as tea stem insects, so it is urgent to develop a new and non-destructive detection technology and method. In order to solve the problem of overlapping and covering foreign bodies in tea leaves, a scheme of electromagnetic vibration feeding assisted near-infrared spectroscopy(NIRS), and fluorescence image was proposed to online detect endogenous foreign bodies of tea stalks and insects.A total of 600 NIRSranging from 600 to 1 050 nm, and 65 channel images including R, G, B and N were collected by electromagnetic vibration-assisted NIRS and fluorescence imaging system. Among them, 451 spectra were used to develop the model, and the remaining 149 spectra were used to evaluate model performance. The effects of different correction methods such as detrending, multiplicative scatter correction (MSC), standard normal variate transformation (SNV), variable sorting for normalization(VSN), adaptive iteratively reweighted penalized least squares(airPLS), alternative least squares(ALS),optical path length estimation and correction (OPLEC) were compared. OPLEC could eliminate the scattering effect better, and the correct recognition rate of the partial least squares discriminant analysis (PLS-DA) model of NIRS increased from 78% to 85%. The results showed that the calibration method of OPLEC combined with the PLS-DA model could- detect foreign bodies in tea more accurately.Compared with the accurate measurement of NIRS, imaging technologyprovided a wider range of detection means. Sixty-five clear blue (B) channel images were analyzed. Using threshold segmentation by maximum interclass variance method, inversing operation, median filtering, connected component labeling and feature extraction, we extracted four feature variables of long axis length, short axis length, short axis ratio and eccentricity, a total of 355 objects of interest.The linear discriminant analysis (LDA) model was established with 267 interesting targets, and 88 interested targets not involved in modeling were used to evaluate the model’s prediction ability. The correct recognition rate reached 64%.The experimental results show that electromagnetic vibration feeding assisted NIRS and fluorescence image is feasible for online detection of tea stalk and foreign insect bodies, providing a low-cost solution for online detection of organic foreign bodies in food.
		2023 Vol. 43 (01): 100-106 [Abstract] ( 86 ) RICH HTML PDF (3632 KB) ( 34 )

	107	*Detection Method of Freshness of Penaeus Vannamei* Based on Hyperspectral**
		ZHU Chen-guang¹, LIU Ya-jun², LI Xin-xing^{1, 3}, GONG Wei-wei^4*, GUO Wei¹
		DOI: 10.3964/j.issn.1000-0593(2023)01-0107-04
		In this study, Penaeus vannamei was taken as the research object to explore an efficient, rapid and non-destructive freshness detection method. Total volatile basic nitrogen（TVB-N）is an important chemical index to judge the freshness of shrimp. However, the traditional method is time-consuming and labor-consuming, which limits the real-time detection of large quantities. In recent years, hyperspectral technology has been an analysis technology integrating image and spectral information. Each pixel in the hyperspectral image contains the spectral information of the whole band. This technology has become a technology of meat freshness detection. This study collected 860~1 700 nm hyperspectral data of Penaeus vannamei samples for 8 consecutive days. After removing the abnormal samples, 150 groups of test samples were determined. We collected 254-dimensional spectral data in each group. The original hyperspectral image was corrected in black and white, and ENVI software extracted the spectral data from the hyperspectral image. We ensured the corresponding relationship between the extracted spectral data and the TVB-N index. The average spectrum of the ROI is calculated to obtain the spectral data matrix, which is converted into ASCII code and saved. At the same time, the true value of TVB-N was obtained by the Kjeldahl method. In order to reduce the interference of water content of environment and shrimp surface and effectively eliminate the irrelevant information and noise, this study used a multiple scattering correction algorithms to preprocess the shrimp hyperspectral and selected seven sensitive bands. Finally, a quantitative prediction model of TVB-N of Penaeus vannamei was established based on 120 training set samples and 30 validation set samples. We compared the model of BPNN, RBFNN and PCA. The r and NRMSE of the BPNN model were 0.902 1 and 0.214 0, the RBFNN model were 0.868 3 and 0.223 0, PCR model were 0.757 6 and 0.390 0, respectively. The results showed that the MSC-BPNN model had the best prediction effect, and there is a close correlation between hyperspectral reflectance and freshness of Penaeus vannamei. This paper supports the freshness detection of shrimp based on spectral technology.
		2023 Vol. 43 (01): 107-110 [Abstract] ( 114 ) RICH HTML PDF (1809 KB) ( 137 )

	111	A Photoelastic Modulator Based MSE Spectroscopic Diagnostic on EAST
		LI Yi-chao^{1, 2}, FU Jia^1, LÜ Bo^1, HUANG Yao¹, QIAN Jin-ping¹, LU Zheng-ping¹, FU Sheng-yu¹, LI Jian-kang¹, WEI Yong-qing³, LIU Dong-mei⁴, XIAO Bing-jia¹
		DOI: 10.3964/j.issn.1000-0593(2023)01-0111-05
		Current density distribution is a crucial parameter in plasma physics, which plays an important role in plasma simulation, advanced operation mode development, current drive, confinement, and transport. The stark spectrum caused by injecting a neutral beam into plasma contains σ component, and π component. When viewed transversely to the electric field, the polarization direction of the σ component is perpendicular to the direction of the equivalent electric field, and the polarization direction of π component is parallel to the equivalent electric field, the distribution of current plasma density can be deduced by measuring polarization direction of the splitting spectrum.The polarization detection system based on a dual photoelastic modulator system has the unique advantages of high detection accuracy and quick time response, which is very suitable for measuring current density distribution under the rapid variation of plasma current. An external driving source causes the elastic deformation of birefringent crystals, and the refractive index changes periodically, then the modulated light intensity changes are formed through the polarizer. The polarization detection system of MSE diagnostic consists of two Photoelastic Modulators (PEM) and a polarizer. By detecting the ratio of modulation intensities at different modulation frequencies, the polarization direction of the splitting spectrum can be obtained quickly and accurately, and then the plasma current density distribution can be obtained. This paper introduces the polarization detection system of MSE diagnosis on Experimental Advanced Superconducting Tokamak (EAST), The off-line test and calibration of the system were completed, and the preliminary result of plasma current density distribution was obtained in experiments.
		2023 Vol. 43 (01): 111-115 [Abstract] ( 96 ) RICH HTML PDF (2260 KB) ( 26 )

	116	Research on Parameter Optimization of Apple Sugar Model Based on Near-Infrared On-Line Device
		JIANG Xiao-gang¹, ZHU Ming-wang¹, YAO Jin-liang¹, LI Bin¹, LIAO Jun¹, LIU Yan-de^1*, ZHANG Jian-yi², JING Han-song²
		DOI: 10.3964/j.issn.1000-0593(2023)01-0116-06
		Soluble solids content is one of the leading evaluation indicators for internal apple quality. NIR spectroscopy is the first choice for predicting apple soluble solids. Optimizing the parameters of near-infrared spectroscopy collection devices can improve the model’s performance. In this paper, the near-infrared spectrum (350~1 150 nm) of apples was collected by the dynamic online equipment independently developed by our research group, and the effects of different parameters (movement speed, integration time, and light intensity) on the apple quality prediction model by near-infrared spectrum were studied, the parameters of the dynamic online equipment were optimized. The 210 Fuji apples were divided into two batches. The first batch of 90 apple samples was divided into a modeling set and a prediction set by the K-S algorithm, which was used to study the effect of the online prediction model on the solid soluble content of apples with different movement speeds and different integration times. At two moving speeds of 0.3 and 0.5 m·s^-1, multiple scattering correction (MSC) and wavelet transform (WT) are used to preprocess the collected spectra, and the SSC model is built for the spectra with different moving speeds. The results show that the prediction model built with amoving speed of 0.5 m·s^-1 performs better. Among the four different integration times, the best performance of the model built by SNV preprocessing was achieved at an integration time of 120 ms. The second batch of 120 apples was divided into modeling and prediction sets by the K-S algorithm. The influence of different light intensities on the apple’s SSC prediction model was studied using device parameters with a moving speed of 0.5 m·s^-1 and integration time of 120ms. The results showed that when the light intensity was 4.5 A, the collected spectrum changed significantly compared with other light intensity groups, and the peaks at 640 and 800 nm of the spectrum disappeared. When the light intensity is 6.5A, the model after SNV pretreatment has the best performance. Competitive Adaptive Reweighting Algorithm (CARS) and Successive Projections Algorithm (SPA) were used to screen the wavelength of the collected spectral data to establish the apple SSC model. The results show that the model-based on CARS-PLS has good performance and the correlation coefficient and root mean square error of its prediction set are 0.991 and 0.149, respectively. At the same time, the model is simplified, and the stability of the model is improved. The research shows that parameter optimization of dynamic online equipment is helpful in improving the prediction accuracy of the apple model. This research is beneficial in providing technical support for online apple quality sorting.
		2023 Vol. 43 (01): 116-121 [Abstract] ( 125 ) RICH HTML PDF (2457 KB) ( 152 )

	122	Coordination Interaction of DSAZn With Quercetin and High Sensitivity Detection of Quercetin
		ZHAI Yan-ke¹, PAN Yi-xing¹, XIANG Hao¹, XU Li^1*, ZHU Ze-ce², LEI Mi¹
		DOI: 10.3964/j.issn.1000-0593(2023)01-0122-07
		Quercetin is a natural flavonoid compound that is used for the prevention and treatment of hypertension, hyperlipidemia, cardiovascular disease, cancer, etc. Therefore, quantitative detection of quercetin was particularly important in biochemistry and clinical medicine. A highly selective and sensitive detection methodfor quercetin with AIE (aggregation-induced luminescence phenomenon) fluorescent molecules was proposed, which identified the target molecule quercetin through coordination interaction with excited state electron transfer. The fluorescence of DSAZn in PBS buffer at pH 7.0 with adding five drug molecules (quercetin, icaritin, isorhamnetin, rutin and dopamine) was studied.The fluorescence emission spectra of 435~680 nm were scanned by a fluorescence spectrophotometer with an excitation wavelength of 415 nm. The ultraviolet absorption spectrum of 250~750 nm was scanned by ultraviolet spectrophotometer, which showed that the traditional Chinese medicine molecule quercetin could form a complex with the AIE fluorescent probe and thusstaticallyquenchedthe fluorescence of the AIE probe. Fluorescence detection showed that the quenching effect of five drug molecules on the fluorescent probe was significantly different. The binding constant of quercetin and DSAZn was 1.34×10⁷ L·mol^-1, which was an order of magnitude higher than that of the other four drug molecules, exhibitinga good selectivity for quercetin. The detection limit of quercetin was 3.07 nmol·L^-1, which was lower than the value reported in many kinds of literature, exhibiting a high sensitivity for quercetin. The titrationcurve equation of quercetin to DSAZn based on the fluorescence titration spectrum was: y=0.013 4x-0.294 82. The linear range of quercetin concentration was 0～5 μmol·L^-1 with a linear correlation coefficient of 0.994 3. Thus, a highly selective and sensitivedetection method for quercetin by AIE-type fluorescent molecules was constructed. This method was simply operated and repeatable, which provided a new research idea for detecting drugs with similar structures.
		2023 Vol. 43 (01): 122-128 [Abstract] ( 95 ) RICH HTML PDF (5630 KB) ( 38 )

	129	Study on the Radiobiological Effects of Low-Dose X-Ray on Human Neuroblastoma Cells by Raman Spectroscopy
		CHEN Shan¹, LIN Lan², CHEN Jun³, LIU Ying^1*
		DOI: 10.3964/j.issn.1000-0593(2023)01-0129-04
		The biological effects induced by low-dose ionizing radiation are complex and diverse. Radiation biomarkers and detection techniques often limit the studies. In this paper, Raman spectroscopy was applied to study the biological effects of low-dose radiation. Raman spectroscopy at 10 mW and 532 nm were used to analyze human neuroblastoma cells irradiated by X-ray at 100, 200 and 500 mGy. The DNA related Raman characteristic spectroscopic peaks of purine nucleotides (722～728, 1 572～1 581 cm^-1, etc. ) and pyrimidine nucleotides (770～785 cm^-1, etc.) were changed by ionizing radiation, indicating that low-dose X-ray irradiation caused changes in cell DNA level. Flow cytometry was used to analyze the cell cycle of human neuroblastoma cells cultured for 6 hours after irradiation under the same conditions. All three doses of X-ray ionizing radiation caused cell stagnation in the G2 phase of the cell cycle, which also suggested that ionizing radiation caused the increase of DNA level. Analysis of cell migration at 20 h after irradiation by scratch assay showed that human neuroblastoma cells exposed to all three doses of ionizing radiation showed reduced levels of migration compared to control cells not exposed to X-ray. Raman spectroscopy showed that low dose X-ray ionizing radiation induced changes in the DNA level of human neuroblastoma cells, and that was consistent with the cell cycle analysis and migration analysis, but the detection time was much earlier. Raman spectroscopy can be used to detect and monitor the early biological effects of low-dose radiation.
		2023 Vol. 43 (01): 129-132 [Abstract] ( 100 ) RICH HTML PDF (2333 KB) ( 38 )

	133	*Discriminating Flavor Styles via* Data Fusion of NIR and EN**
		WANG Wen-jun¹, SHA Yun-fei¹, WANG Yang-zhong¹, YU Jie¹, LIU Tai-ang², ZHANG Xu-feng³, MENG Xiang-zhou³, GE Jiong^1*
		DOI: 10.3964/j.issn.1000-0593(2023)01-0133-05
		In this study, a qualitative discrimination model was established based on the combined technology of near-infrared (NIR) and electronic nose (EN) to distinguish the light, intermediate and strong flavors of tobacco leaves. The results showed little difference in the accuracy of the three models, all of which were more than 89.00%. However, the prediction accuracy of the combined model for intermediate flavor and strong flavor was 82.67% and 80.00%, respectively, which were significantly higher than those by NIR (72.41% and 73.33%) and EN (68.97% and 53.33%). The reason may be that EN was more sensitive to aroma components affecting intermediate flavor and strong flavor, and captured more information. The new information as a beneficial supplement to NIR data and can be used to establish a model with higher accuracy for tobacco flavor classification. In addition, based on the same fusion data, this study compared the modeling and prediction accuracy of different data mining algorithms. The results showed that the modeling accuracy of the artificial neural network (99.07%) was higher than that of the support vector machine (96.26%). However, the prediction accuracy of the artificial neural network (65.00%) was significantly lower than that of the support vector machine (83.75%), which verified that the support vector machine could reduce overfitting in the modeling process. This study can support the rapid identification of tobacco flavor style, and the further development of this technology will strive to provide an auxiliary identification method for professional tobacco evaluators.
		2023 Vol. 43 (01): 133-137 [Abstract] ( 124 ) RICH HTML PDF (1600 KB) ( 40 )

	138	*Research on Origin Traceability of Rhizoma Dioscoreae* Based on LIBS**
		CAI Yu^{1, 2}, ZHAO Zhi-fang³, GUO Lian-bo⁴, CHEN Yun-zhong^{1, 2*}, JIANG Qiong⁴, LIU Si-min^{1, 2}, ZHANG Cong-zi⁴, KOU Wei-ping⁵, HU Xiu-juan⁵, DENG Fan⁶, HUANG Wei-hua⁷
		DOI: 10.3964/j.issn.1000-0593(2023)01-0138-07
		Rhizoma Dioscoreae contains polysaccharides, polyphenols, saponins, mucins and vitamin C, which have anti-tumor, anti-oxidant, anti-inflammatory, hypoglycemic, and hypolipidemic effects. Due to the differences in growth conditions of Rhizoma Dioscoreae from different origins, resulting in significantly different contents of medicinal ingredients, combined with unique processing technology, which in turn lead to large differences in market prices, it is crucial to identify the origin of Rhizoma Dioscoreae Tablets. In order to trace the origin of Rhizoma Dioscoreae Tablets, this paper proposed a Multiplicative signal correction-improved genetic algorithm-support vector machine (MSC-IGA-SVM) model based on Laser-induced breakdown spectroscopy (LIBS) technique for accurate identification of Rhizoma Dioscoreae origin. In the paper, LIBS experiments were conducted using eight Rhizoma Dioscoreae Tablets of different origins. The Rhizoma Dioscoreae Tablets of eight origins were ground and sieved to make powder pressed tablets. The recognition results of the spectra were compared by collecting LIBS spectra of Rhizoma Dioscoreae Tablets using a single classifier and a model using spectral preprocessing, feature extraction and pattern recognition algorithms, respectively. In the research, the spectral signals were divided into training and test sets in the ratio of 2∶1, and the test set accuracy of the K-Nearest Neighbor (KNN) model using five cross-validations was used as an evaluation index for the optimization of preprocessing parameters. The overall trend of the average spectra of all herbs was consistent. The contained spectral peaks were the same, but the peak intensities varied due to different origins, and the enrichment ability of some metal elements (K, Na, Ca, Mg, Al) was greater for Rhizoma Dioscoreae growed in the Dao-di Areas than for those not growed in the Dao-di Areas, among which, the peak of the characteristic spectral line of element K (769.90 nm) was the highest, i. e., the Rhizoma Dioscoreae Tablets contained the most element K. Related studies showed that the root of Rhizoma Dioscoreae has the strongest enrichment capacity for element K. Thirty-five key spectral lines were selected for analysis. Improved Genetic Algorithm (IGA) could discriminate the nonlinear relationships in the spectra more clearly than Principal Component Analysis (PCA) in the case of many identification species and difficult identification while being less affected by noise. The MSC-IGA-SVM model had the best origin traceability. The accuracy of the MSC-IGA-SVM model was 96.9% for the cross-validation set, and the accuracy of the test set was 97.32%, which was 0.87% higher than the best model Support Vector Machine (SVM) built directly using the original signal (96.43%) for the test set. Meanwhile, the MSC-IGA-SVM model reduced the dimensionality of the input variables by 99.93%. The results showed that the origin of Rhizoma Dioscoreae Tablets could be traced by the LIBS technique combined with the MSC-IGA-SVM model quickly and accurately.
		2023 Vol. 43 (01): 138-144 [Abstract] ( 122 ) RICH HTML PDF (3684 KB) ( 106 )

	145	Rapid Qualitative Analysis of Synthetic Cannabinoids by Raman Spectroscopy
		HU Shuang¹, LIU Cui-mei^2*, JIA Wei², HUA Zhen-dong²
		DOI: 10.3964/j.issn.1000-0593(2023)01-0145-06
		In July 2021, China imposed a class-wide ban on synthetic cannabinoids, with the definition of synthetic cannabinoids by seven core structures. It is an urgent need for the drug law-enforcing departments and relevant technicians to conduct in-field rapid qualitative analysis of suspected synthetic cannabinoid samples. This study investigated the overall discriminant ability of Raman spectroscopy for synthetic cannabinoids, compared four handheld Raman spectrometers, and discussed the possible reasons that restricted the wideapplication of Raman spectroscopy. ProTT-EZRaman-A7 portable Raman spectrometer, with the overall performance between desktop Raman and handheld Raman, was selected to collect the Raman spectra of 90 synthetic cannabinoid reference substances. Then a general Raman spectrum library with 90 synthetic cannabinoids was established using KnowItAll software, which was compatible with various original spectrum formats. The analysis of 90 synthetic cannabinoid Raman spectra found that when there was no fluorescence interference, Raman spectroscopy could distinguish all synthetic cannabinoids, but showed a low discriminant ability for some structural analogues, especially those with a difference of a methyl or a halogen atom. The performance of different Raman spectrometers varied greatly, so in order to investigate the reasons for that performance differences, four handheld Raman spectrometers were selected to analyze 120 seized synthetic cannabinoid samples. Then the library search was performed using the KnowItAll software and the established general Raman spectrum library. The correct matching rates of the four Raman spectrometers were 71.7%, 68.3%, 46.7%, and 24.2%, respectively. The difference in background fluorescence reduction effect and resolution attributed to that result. The portable Raman spectrometer was simple, fast, and can be used for in-field testing. However, Raman spectroscopy can only be used for preliminary qualitative screening considering the unknown purity of the seized samples, the possible fluorescence interference, the performance of different Raman spectrometers, and the spectral library completeness. This study provided guidance for forensic science laboratories and relevant technicians to apply the Raman spectrum results correctly.
		2023 Vol. 43 (01): 145-150 [Abstract] ( 94 ) RICH HTML PDF (3203 KB) ( 60 )

	151	Fast Resolution Algorithm for Overlapping Peaks Based on Multi-Peak Synergy and Pure Element Characteristic Peak Area Normalization
		CHEN Ji-wen, YANG Zhen, ZHANG Shuai, CUI En-di, LI Ming^*
		DOI: 10.3964/j.issn.1000-0593(2023)01-0151-05
		Due to the mutual interference of characteristic peaks between elements and affected by the energy resolution of the experimental instrument, overlapping peaks will be formed when the characteristic peaks of multiple elements have similar peak positions and widen broadening. Taking the overlapping peaks with low resolution and high-resolution requirements as the research object, a fast resolution algorithm for overlapping peaks based on multi-peak synergy and pure element characteristic peak area was proposed, and combined with the actual X-ray fluorescence spectrum to verify the new method. Taking the X-ray fluorescence spectrum of dysprosium-ferrum alloy as an example, under the experimental conditions of this paper, the separation degree of the overlapping peaks formed by the characteristic peaks of dysprosium L_α and iron K_α is about 0.273 5. The lower L_β characteristic peaks of dysprosium and iron K_β characteristic peaks. First, configure the dysprosium standard solution in the concentration range (7.8～8.2 mg·mL^-1) and the iron standard solution in the concentration range (1.8～2.2 mg·mL^-1) to measure and obtain pure element spectra and calculate the area respectively. Normalized and averaged to obtain normalized characteristic peaks of dysprosium L_α and iron K_α peak. Then, use dysprosium and iron standard solutions to mix 20 groups of sample solutions with a mass percentage of iron elements ranging from 19.1% to 21% and a step of 0.1% for measurement. Since the overlapping peak part is only composed of the dysprosium L_α peak and the iron K_α peak, the weight of the dysprosium L_α peak plus the iron K_α peak weight in the overlapping peak is set to 1, using the dysprosium L_α peak and the iron peak. The normalized characteristic peaks of the K_α peaks are fitted for overlapping peaks. The approximate weight range of iron element is determined by the characteristic peaks of dysprosium L_β and iron K_β, and the weight value optimization is carried out with the particle swarm optimization algorithm to complete the decomposition of overlapping peaks. By fitting the iron element weight obtained by the optimization solution with the iron element mass percentage of the actual sample solution, a regression line from the iron element weight in the dysprosium iron sample solution to the actual iron element mass percentage is obtained. Finally, the actual sample experiment was carried out, and the iron content detected by the method in this paper was compared with the reference value of iron content determined by the national standard potassium dichromate volumetric method. The results show that the rapid resolution algorithm of overlapping peaks based on multi-peak synergy and normalization of characteristic peak areas of pure elements can resolve overlapping peaks with low resolution and high resolution requirements.
		2023 Vol. 43 (01): 151-155 [Abstract] ( 141 ) RICH HTML PDF (2399 KB) ( 68 )

	156	*Fast Evaluation of Freshness in Crayfish (Prokaryophyllus clarkii) Cased on Near-Infrared Spectroscopy*
		WANG Chao¹, LIU Yan^1*, XIA Zhen-zhen², WANG Qiao¹, DUAN Shuo¹
		DOI: 10.3964/j.issn.1000-0593(2023)01-0156-06
		Crayfish is one of the most popular freshwater products in China. The industrial chain of crayfish has rapidly developed and produced gorgeous economic benefits. Easy to be putrid during the logistics transportation, the freshness of crayfish and related products must be monitored and has paid much attention in recent years. If the putrid crayfish cannot be detected in time, food safety accidents may happen, and the whole industrial chain of crayfish would be destroyed. The total volatile basic nitrogen (TVBN) is the common index of freshness for aquatic products and can be used to evaluate the freshness of crayfish. The traditional analytical methods for TVBN are accurate but complex, time-consuming and environmentally hazardous. Developing novel, fast and stable methods are inevitable for the freshness evaluation of crayfish with large scale. Near-infrared spectroscopy (NIR) is a fast, non-destructive and environmentally friendly analytical technique widely used in many fields. In this study, a method for monitoring the freshness of crayfish by near-infrared spectroscopy combined with chemometrics was proposed. The TVBN were adopted as the freshness index and the quantitative models were built by partial least squares (PLS). The spectral pretreatment and variable selection methods were adopted to improve the models further. For the edible part of the crayfish, reasonable validation results can be obtained by using the optimized models. The combination of 1st and (MC-UVE) seems to have the better optimization results. For total volatile basic nitrogen (TVBN), the root means square error of prediction (RMSEP) and correlation coefficient (r) of the crayfish tails were 1.626 and 0.950.
		2023 Vol. 43 (01): 156-161 [Abstract] ( 105 ) RICH HTML PDF (1909 KB) ( 115 )

	162	Quantitative Characterization of Wheat Starch Retrogradation by Combining 2D-COS and Spectral Fusion
		AN Huan-jiong¹, ZHAI Chen², MA Qian-yun¹, ZHANG Fan¹, WANG Shu-ya², SUN Jian-feng¹, WANG Wen-xiu^1*
		DOI: 10.3964/j.issn.1000-0593(2023)01-0162-07
		Retrogradation is an important physicochemical property of starch during processing, transportation and storage. Rapid detection of retrogradation is of great significance to starch products’ quality and shelf life. In order to investigate the feasibility of selecting the characteristic variables of retrograde starch by two-dimensional correlation spectroscopy (2D-COS), spectral fusion technology and 2D-COS was combined to quantitatively characterize the retrogradation characteristics of wheat starch in this study. First, wheat starch’s crystallinity and retrogradation degree at different retrograde times were measured. The retrograde properties of starch were characterized by crystal content in the starch system and resistance to amylase hydrolysis. Then, the samples’ near-infrared and mid-infrared spectral data were collected respectively. After spectral pretreatment, prediction models based on near-infrared, mid-infrared, -and fusion spectra were established using partial least squares analysis. On this basis, the retrogradation day was used as the external disturbance. Starch spectra of 0, 1, 2, 3, 5, 7, 10, 14, 21 and 35 days were selected for 2D-COS analysis. By analyzing the synchronization and autocorrelation spectrum, 13 and 11 feature variables related to starch retrogradation characteristics were identified from near-infrared and mid-infrared spectra, respectively. Finally, prediction models for retrogradation degree and crystallinity were established based on these variables. The results show that the models based on full-spectra yielded better prediction performance after spectral fusion, with relative percent deviation (RPD) increasing from 1.203 4 and 2.069 0 to 3.980 9 and from 2.594 0 and 2.109 9 to 4.576 3 for crystallinity and retrogradation degree. Using the feature spectra obtained by 2D-COS analysis, the RPD values for the crystallinity model and retrogradation degree model increased to 8.095 9 and 14.183 6. 2D-COS can improve spectral resolution and obtain more chemical structure information than the model based on Competitive Adaptive Reweighted Sampling. Therefore, the spectral fusion technology combined with 2D-COS model has better results. The results show that it is feasible to use the 2D-COS to identify the characteristic wavelengths for starch retrogradation properties, which provides a new idea for the characteristic variables optimization of fusion spectra. Spectral fusion technology combined with 2D-COS can realize the rapid detection of starch retrograde, which provides a method for rapidly detecting starch food quality and shelf life.
		2023 Vol. 43 (01): 162-168 [Abstract] ( 93 ) RICH HTML PDF (4338 KB) ( 67 )

	169	Study on Sample Preparation Method of Plant Powder Samples for Total Reflection X-Ray Fluorescence Analysis
		JIA Wen-bao¹, LI Jun¹, ZHANG Xin-lei¹, YANG Xiao-yan², SHAO Jin-fa³, CHEN Qi-yan¹, SHAN Qing^1*LING Yong-sheng¹, HEI Da-qian⁴
		DOI: 10.3964/j.issn.1000-0593(2023)01-0169-06
		The knowledge of tailing slurry’s heavy metal elementalry grade is important in mineral flotation processing. It can improve mineral utilization andreduce environmental pollution. X-ray fluorescence spectroscopy is an effective technology for determining heavy metal elements. Compton scattering internal standard is commonly used in X-ray fluorescence spectroscopy quantitation for geological samples. However, for thin film samples, the Compton scattering peak will be affected by the filter. However, it wasn’t easy to measure the Compton scattering intensity of the sample directly because it was tightly attached to the filter.In this paper, the filter’s influence on the sample’s Compton scattering intensity was discussed for the tailing slurry’s thin film sample after filtration, and the Compton scattering intensity of the thin film sample was corrected. The experimental results show that the intensity of the Compton scattering peak increases linearly with the increase of the thickness of the polypropylene filterin the thickness range of 0.34～3.06 mm.Therefore, the linear relationship between the total Compton scattering intensity obtained by the detector and the filter thickness was established, and the scattering peak intensity was the true Compton scattering peak intensity of the sample when the filter thickness was 0. Afterwards, Monte Carlo simulated the Compton scattering of samples with no filter and samples with different filter thicknesses. The results showed that the corrected Compton scattering intensity was the same as that of the sample without filter, with a relative deviation of only 0.41%. The ratio of the corrected Compton scattering intensity to the uncorrected Compton scattering peak were 91.31% and 89.91%, compared through experiments and simulations when the thickness ofthe polypropylene filter is 0.34 mm, which has a good consistency. Finally, the standard curves of elements were established and corrected by the uncorrected and corrected Compton scattering internal standards after six standard materials were measured experimentally.Quantitative analysis of the Cu, Zn and Pb elements in the two tailings slurry showed that the uncorrected Compton internal standard correction increased the relative deviation of some elementsby 3.18% to 9.00% compared with the ICP-OES results before correction. However, the relative error between the quantitative result of the corrected Compton internal standard method. Moreover, the ICP-OES result was between 1.14% and 11.15%, which was reduced by 0.30% to 8.97% before the correction.
		2023 Vol. 43 (01): 169-174 [Abstract] ( 135 ) RICH HTML PDF (2816 KB) ( 224 )

	175	Study of the Painted Warrior Figurines Excavated in the Tomb of the Sutong Family (Tang Dynasty) by Spectroscopic Techniques
		LIANG Jia-xiang¹, WANG Tian^1*, ZHANG Ya-xu², WANG Fen¹, LI Qiang³, LUO Hong-jie³, ZHAO Xi-chen², ZHU Jian-feng¹
		DOI: 10.3964/j.issn.1000-0593(2023)01-0175-08
		Warrior Figurines is an important God and ghost Figurine in the tomb of the Tang Dynasty. It is the tomb of dignitaries and dignitaries in the Central Plains. In order to explore the color elements and color painting technology of the color painting pigments of the terracotta warriors，X-ray fluorescence spectroscopy is used to analyze the elemental composition of the painted area of the king terracotta figures unearthed from the pits KTJ-2019-M₂ and KTJ-2019-M₃ of the Sutong family tomb in Weicheng District, Xianyang City, Shaanxi Province. The results show that the gold patch on the surface of the terracotta figures is mainly gold foil (Au); the constituent elements of red pigment are Hg, S and a small amount of Pb and P; The constituent elements of blue pigment and green pigment are Cu, and the constituent elements of white pigment are P, S and Pb. Raman spectroscopy was used to identify the phase of color painting pigment layer. The identification results showed that the phase of red pigment was the mixed pigment of cinnabar (HgS) and Pb₃O₄; The phase of blue pigment is Azurite; The phase of green pigment is Malachite green; White pigments may be Cerussite. Further, XRF surface scanning technology was used to analyze the color painting process of the heavenly king figurines, and the element positions in the gold, red, blue and green areas were analyzed. It was found that the color painting of M₂-1，M₃-1，M₃-2，M₃-3，M₃-4 in the ceramic block samples used a single-layer process. The M₂-2 sample color is based on Cerussite and then applied with a layer of cinnabar and minimum mixed pigment, that is, the double-layer color painting process is used. By comparing the test results of the color painting layer of the king terracotta figures of the Sutong Family Tomb of the Tang Dynasty with the color painting cultural relics unearthed in various regions of the Tang Dynasty, it is found that the color painting pigments used in the Sutong Family Tomb of the Tang Dynasty are consistent with the color painting of the color painting cultural relics unearthed in different regions of the same period. The element distribution characteristics of the pigment layer can be obtained intuitively using the surface scanning technology of the X-ray fluorescence spectrum, which can provide reliable scientific support and useful reference for the protection and restoration of the terracotta warriors.
		2023 Vol. 43 (01): 175-182 [Abstract] ( 142 ) RICH HTML PDF (15252 KB) ( 50 )

	183	Classification and Recognition of Lilies Based on Raman Spectroscopy and Machine Learning
		WANG Zhi-xin, WANG Hui-hui, ZHANG Wen-bo, WANG Zhong, LI Yue-e^*
		DOI: 10.3964/j.issn.1000-0593(2023)01-0183-07
		Lily bulbs, the underground metamorphic stems composed of thick scales grown by perennial herbaceous bulbous plants of the lily family Liliaceae, a typical medicinal and edible homologous crop. It is rich in nutrients and has anti-tumor, antidepressant, hypoglycemia, and improves immune functions. The market prices of lily bulbs from different origins are quite different. The traditional evaluation methods that rely on artificial experience and sensory are highly subjective and have poor certainty, making it difficult to be widely used in modern production links. Advanced detection methods based on chemical inspection methods are time-consuming and expensive and it are difficult to meet the requirements for origin identification. Raman spectroscopy is a vibration spectrum based on inelastic scattering, which can achieve fast and accurate non-destructive testing. Combining Raman spectroscopy with machine learning algorithms, a classification model of the three most widely distributed lily bulbs in China (Lanzhou lily, Yixing lily and Longya lily) was established. Observing the characteristic peaks of 479, 870, 942 and 1 606 cm^-1 on the matrix spectrum, a non-destructive testing method based on the component content of Raman spectroscopy to determine the place of origin and evaluate the quality of lily bulbs is proposed. First, the traditional method is used to collect the spectrum of the lily bulb sample. After the spectral data is preprocessed, the artificial prior method is used to extract the representative substance of the lily bulb and determine the characteristic peaks. Then the principal component analysis and the t-distribution random neighborhood embedding method are used to reduce the dimensionality. Extract spectral data features. The data features obtained above are applied to support vector machines, decision trees and random forest algorithms. The experimental results show that these classification models all show ideal classification accuracy on the same test set. Among them, the model’s accuracy based on principal component analysis and decision tree algorithm reached 91.7%。The model’s accuracy based on t-distribution random neighborhood embedding and support vector machine is 93.7%, and the accuracy rate of the model combining the principal component analysis and random forest algorithm is as high as 95.8%. In summary, this method can provide on-site rapid identification and identification of the origin of lily bulbs, improve the accuracy of the quality assessment in the modern production process, and provide a reference for the identification of the origin of modern production and the quality analysis of lily bulbs.
		2023 Vol. 43 (01): 183-189 [Abstract] ( 144 ) RICH HTML PDF (3880 KB) ( 236 )

	190	The Microstructure of “Iron Spot” on Blue-and-White Porcelain From Jingdezhen Imperial Kiln in Yongle and Xuande Period of Ming Dynasty
		WANG Wen-xuan¹, WEN Rui^1*, ZHANG Yue², JIANG Jian-xin³
		DOI: 10.3964/j.issn.1000-0593(2023)01-0190-08
		“Iron spot” refers to the black, cyan and brown spots with metallic tin light condensed from the drawing lines on the blue and white porcelain. As a typical identification feature of blue-and-white porcelain of the early Ming dynasty, it has long been considered related to the use of imported cobalt with high-Fe and low-Mn. Although there have been sporadic reports over the years, due to factors such as submicron crystal size, glaze in homogeneity, element doping, and crystal segregation, its morphology and structure have not been fully studied, resulting in the color mechanism of “iron spot” unclear, and the view that “iron spot” as an identification standard for imported cobalt pigment being questioned. Combined with previous research, we found that Raman spectroscopy and scanning electron microscope equipped with energy dispersive spectrometer have great advantages in analysing ancient ceramic microcrystal structures. In order to further explore the composition and structure characteristics of “iron spot”, the ultra-depth three-dimensional video microscope, scanning electron microscope with energy dispersive spectrometer, and Raman spectrometer was used to analyze the crystal microstructure of “iron spot” in five blue-and-white porcelain samples of Jingdezhen imperial kiln in the Yongle and Xuande period of Ming dynasty. The composition of the samples’ white glaze area, blue color area and “iron spot” area was tested by laser ablation inductively coupled plasma emission spectroscopy. The microscopic observation results show that the diversity of crystallographic morphology and distribution in the “iron spot” area of different samples is the main reason for many visual perceptions: the octahedron and its massive aggregate crystals precipitated on the glaze surface present a point-like flashing visual perception; the dendrites that are oriented in parallel and well-developed are prone to the phenomenon of “tin light”; the frosted visual perception is caused by the close arrangement of dendritic and snowflake-like crystals; The excessive development of anorthite forms the raised brown spot to form a network structure. In terms of microstructure, the dendrites in the Yongle period are mainly composed of Mg²⁺-doped CoFe₂O₄-Fe₃O₄ solid solution, while crystals in the Xuande period are mainly MnFe₂O₄-Mn₃O₄ solid solution doped with Mg²⁺ and Co²⁺, and associated with reticulated anorthite. The above results show that an “iron spot” can be formed on blue-and-white porcelain fired with imported or domestic cobalt pigment. Since the crystals formed are all cubic inverse spinel structures, they have a certain similarity in macroscopic morphology. To sum up, this study clarifies the microstructure and composition characteristics of “iron spot” on blue-and-white porcelain from Jingdezhen imperial kiln in the Yongle and Xuande period of the Ming Dynasty. It reveals the coloring mechanism of “iron spot”, which provides a certain scientific basis for the identification of blue-and-white porcelain in Jingdezhen imperial kilns, and also provide some reference for the application of Raman spectroscopy and scanning electron microscopy spectroscopy equipped with energy dispersive spectrometer in the analysis of complex microcrystalline structures of ancient ceramics.
		2023 Vol. 43 (01): 190-197 [Abstract] ( 231 ) RICH HTML PDF (4443 KB) ( 60 )

	198	Comparative Analysis of GF-1 and GF-6 WFV Images in Suspended Matter Concentration Inversion in Dianchi Lake
		ZHAO Ran¹, YANG Feng-yun^1*, MENG Qing-yan^{2, 3}, KANG Yu-peng^{2, 4}, ZHENG Jia-yuan¹, HU Xin-li², YANG Hang²
		DOI: 10.3964/j.issn.1000-0593(2023)01-0198-08
		Total suspended matter (TSM) is one of the important parameters of water environment assessment and an important index of remote sensing water retrieval. GF-1/WFV and GF-6/WFV are free and open satellite data of the gaofen series, which are widely used in remote sensing monitoring. However, there are few studies on the applicability of the new bands of GF-6/WFV in water quality parameter inversion. This study takes Dianchi Lake in Yunnan province as the research area, based on the testing data synchronization with water transit (or similar) of the phase of GF-1/WFV and GF-6/WFV remote sensing image using statistics analysis method to the same band (blue, green, red and near-infrared) consistency analysis, regression method based on using the experience of the TSM inversion models of the two kinds of data, respectively, The model with GF-6/WFV added bands were compared with the model constructed by GF-1/WFV. The optimal model was applied to six GF-6/WFV images in 2020 to obtain the TSM distribution map of Dianchi Lake. The results show that the correlation coefficients of GF-1/WFV and GF-6/WFV in blue, green, red and near infrared bands are 0.98, 0.98, 0.97 and 0.99, respectively. The apparent reflectance of the two kinds of data is highly consistent. The inversion accuracy of GF-1/WFV difference model “B2+B4-B1” based on blue, green and near-infrared red bands is high, and the root means square error of model inversion is 6.35 mg·L^-1, and the average absolute percentage error is 23.60%. The ratio model “1/B5+B6” constructed by GF-6/WFV based on near-infrared, red-edge 1 and red-edge 2 bands has a high inversion accuracy. Model inversion’s root mean square error (RMSE) is 3.07 mg·L^-1, and the mean absolute percentage error (MAPE) is 20.65%. By comparing the difference model “B1-B4” constructed by GF-1/WFV with “B5-B4” constructed by GF-6/WFV, it is found that the root means square error of the latter is reduced by 2.61 mg·L^-1, and the average absolute percentage is reduced by 32.33%. The experiment shows that the inversion effect of the model with the red-edge band is better than other models. The TSM distribution map of Dianchi Lake in 2020 was obtained using the modeling formula. The TSM in Dianchi Lake varied from 4 to 45 mg·L^-1, with an average value of 18.23 mg·L^-1. The overall spatial distribution showed a trend of heavy distribution in the north and light distribution in the south, and the time distribution of TSM in Dianchi Lake showed an upward and downward trend. This study can not only provide a reference for the sensor band setting of lake water quality monitoring but also provide technical support for water quality remote sensing monitoring by the water resources supervision department of Dianchi Lake.
		2023 Vol. 43 (01): 198-205 [Abstract] ( 98 ) RICH HTML PDF (4709 KB) ( 109 )

	206	Estimation of Arsenic Content in Soil Based on Continuous Wavelet Transform
		WANG Xue-mei^{1, 2}, YUMITI Maiming¹, HUANG Xiao-yu^{1, 2}, LI Rui^{1, 2}, LIU Dong^{1, 2}
		DOI: 10.3964/j.issn.1000-0593(2023)01-0206-07
		Compared with the traditional detection methods, hyperspectral technology has the characteristics of rapid, accurate and low cost in estimating soil heavy metal arsenic content and can dynamically monitor heavy metal arsenic pollution of oasis soils in arid regions. Based on the collection of soil samples from the cultivated layer of the delta oasis of Weigan-Kuqa river in Xinjiang, soil spectral data and heavy metal arsenic content were obtained. Through the four wavelet basis functions bior1.3, db4, gaus4 and mexh, the original spectral reflectance of the soil was subjected to continuous wavelet transformation. The transformed spectral data was correlated with the heavy metal arsenic so that the selected sensitive wavelet coefficients were taken as independent variables, using partial least square regression, support vector machine regression, BP neural network and random forest regression methods to perform hyperspectral inversion of heavy metal arsenic content. The results showed that: (1) The spectral decomposition effect of the four wavelet basis functions at scales 3 to 8 was obviously better than that of other scales, especially the continuous wavelet transform at scales 4 to 6, effectively improved the correlation between the spectral reflectance with soil heavy metal arsenic, and the number of wavelet coefficients passing the significance test increased significantly (p<0.01), and there had a strong correlation in the vicinity of 400~700 nm in visible light and 1 100~1 700 and 2 200~2 400 nm in near-infrared. (2) By comparing the ability of the four wavelet basis functions to identify effective information in the spectral data, it was believed that the wavelet basis functions bior1.3 and mesh were better than db4 and gaus4. Among them, bior1.3 had the best spectral decomposition effect, and gaus4 was relatively weak. Through the spectral transformation of the 5th scale of bior1.3, the number of bands significantly related to soil heavy metal arsenic was the largest, which was 507 (p<0.01). (3) Comparing the inversion results of the four modeling methods, it was found that the SVMR, BPNN and RFR models had stronger estimation capabilities than the PLSR model, and the estimation accuracy of the model was high. After comprehensively analyzing each model’s stability and estimation accuracy, it was concluded that the bior1.3-2⁵-RFR model could be used as the best estimation model for the heavy metal arsenic in the study area. The R² of the training set and the validation set of the model were 0.893 and 0.639 respectively, the RMSE were 1.075 and 1.651 mg·kg^-1, and the RPD were 2.89 and 1.64 respectively, indicating that the model had a better estimation effect and powerful stability. Using appropriate wavelet basis functions to carry out continuous wavelet transform can reduce the white noise in hyperspectral soil data, excavate the effective information in soil spectral data, and provide a strong technical guarantee for accurate estimation of soil heavy metal arsenic content.
		2023 Vol. 43 (01): 206-212 [Abstract] ( 80 ) RICH HTML PDF (4135 KB) ( 83 )

	213	Spectroscopy Characteristics of Emerald From Swat Valley, Pakistan
		BAO Pei-jin¹, CHEN Quan-li^{1, 2*}, WU Yan-han¹, LI Xuan¹, ZHAO An-di¹
		DOI: 10.3964/j.issn.1000-0593(2023)01-0213-07
		With the exhaustion of Colombian emerald mins, the emeralds from Swat vally, Pakistan have gradually dominated the market and the systematic research for emeralds from Swat valley, Pakistan is conducted by using conventional gemological instruments, infrared spectrum, Raman spectrum, UV-Vis-Nir spectrum and Laser Ablation （Microprobe） Inductively Coupled Plasma Mass Spectrometry (LA-ICP-MS). The results show that the color of emerald from swat is dark green to dark bluish green, and the refractive index is 1.589～1.615. There are many kinds and quantities of inclusions in the emeralds from Swat, Pakistan. And the three-phase inclusions are rectangular with a clear boundary and obvious orientation, which is close to the three-phase inclusions in the emeralds from Russia, Zambia and Ethiopia. According to the UV-Vis-Nir spectroscopy and LA-ICP-MS analysis of the same sample with different colors, the UV-Vis-Nir spectroscopy show relatively strong R line absorption of 427, 608, 637 and 679 nm (O light) belonging to Cr in the dark region and absorption of 370 nm belonging to Fe in the o light, and the content of Cr and Fe in this region is relatively high. So emerald bands are caused by different amounts of Cr and Fe. The color of Emerald from Swat, Pakistan, is caused by Cr , and V contributes to color, and Cr/V is very high. According to LA-ICP-MS combined with the infrared spectrum, emerald from Swat , Pakistan belongs to alkali-rich emerald, and the fingerprint region of the infrared spectrum shows the same vibration absorption peak as common emerald. In the middle infrared region of 4 000～2 000 cm^-1, low intensity 3 518 and 3 700 cm^-1 belong to the asymmetric stretching vibration of type Ⅰwater and the other bands caused by water are saturated. The medium strong peak of 3 232 cm^-1 is caused by polymer ion absorption of [Fe₂(OH)₄]²⁺. In the near infrared spectral region of 8 000～5 000 cm^-1, the absorption band at 5 264 cm^-1 belongs to the frequency combination absorption of ν₃+ν₂ of type Ⅰ/Ⅱ water in the direction perpendicular to the C-axis, and the peak of 7 097 cm^-1 is caused by frequency double oscillation of type Ⅱ water. While weak ones of 7 187 and 6 842 cm^-1 are caused by the frequency double oscillation of type Ⅰ water. In the parallel c-axis direction, the absorption band at 5 272 cm^-1 belongs to the synthesis frequency absorption of ν₃+ν₂ of type Ⅰ/Ⅱ water, 7 073 cm^-1 is the synthesis frequency vibration peak of type Ⅰ water, and the peak of 7 185 cm^-1 belongs to the double frequency vibration of type Ⅱ water. In conclusion, the main chromaticity ions of emeralds in Swat Valley of Pakistan are Cr³⁺ and Fe³⁺, and the emeralds’ content of alkali metal ions is high and the emeralds from Swat Valley of Pakistan belong to type Ⅱ water-dominated emeralds.
		2023 Vol. 43 (01): 213-219 [Abstract] ( 144 ) RICH HTML PDF (2818 KB) ( 210 )

	220	The Inversion of Muscovite Content Based on Spectral Absorption Characteristics of Rocks
		ZHAO Jian-ming, YANG Chang-bao, HAN Li-guo^*, ZHU Meng-yao
		DOI: 10.3964/j.issn.1000-0593(2023)01-0220-05
		Rock is composed of various minerals, and there is a close relationship between the reflectance spectral absorption characteristics and mineral content. The spectral absorption characteristics of mineral spectra at specific bands are one of the important indicators for the quantitative estimation of content. This paper takes muscovite as the research object, analysing the rock spectrum’s spectral absorption characteristics near 2.2 μm and muscovite content. Moreover, uses Savitzky-Golay smoothing filter and Continuum Removal method to process the spectral reflectance of rock, and then extracts the spectral absorption characteristic parameters (absorption depth D, absorption width W, absorption area A ), and analyzes the correlation between the absorption characteristics of rock spectrum near 2.2 μm and muscovite content. In this paper, the statistical model was established by a single absorption feature, and the Partial Least Squares (PLS) and Multilayer Perceptron (MLP) models were established by multi-dimensional absorption feature. The muscovite content and spectral absorption characteristic parameters in rocks were analyzed, and a non-linear representation method for predicting mineral content in rocks was proposed. The results show that the spectral absorption characteristics of rock spectrum near 2.2 μm, the correlation between absorption depth and muscovite content among the highest. In the statistical model based on single absorption characteristics, the quadratic curve model has the best fitting effect on the absorption depth. R² is 0.935 0, RMSE is 0.063 0. The absorption depth of the rock spectrum changes with the abundance of muscovite. The higher the muscovite content in rock, the greater the value of rock spectral absorption depth. The PLS model based on multidimensional spectral absorption characteristics was more effective than the MLP model. The R² was 0.947 7 higher than 0.901 2 for MLP, and the RMSE was 0.002 7 lower than 0.005 1 for MLP. On the whole, the multidimensional model is better than the single-dimension model, and the PLS model has the best inversion ability. The model has the characteristics of a small amount of calculation and high precision in predicting muscovite content. Analyzing the spectral absorption characteristics of rocks at the diagnostic characteristics provides a theoretical reference for the quantitative inversion of the content of mineral components. It provides a fast, efficient, and convenient method for the monitoring and evaluating mineral resources.
		2023 Vol. 43 (01): 220-224 [Abstract] ( 107 ) RICH HTML PDF (1610 KB) ( 178 )

	225	Design of Imaging Spectrometer Based on Metasurface
		ZHANG Chun-yu^{1, 2}, ZHOU Jin-song^{1, 2}, HE Xiao-ying¹, JING Juan-juan^{1, 2}, FENG Lei^1*
		DOI: 10.3964/j.issn.1000-0593(2023)01-0225-05
		As an artificially manufactured sub-wavelength structure array plane, metasurface is widely used in many fields because of its light-weight, easy integration, and realization of multiple functions. Traditional spectral imaging systems rely on dispersive components and cumulative phase differences in the optical path to achieve dispersion and focusing of different wavelengths, which cannot meet system integration needs. Unlike traditional optical components relying on the transmission phase accumulated by electromagnetic waves propagating in the medium, the metasurfacerelies on the interface phase mutation for phase control, so a very thin and light optical system can be realized. In this paper, the transmission phase metasurface is studied. The finite difference time domain algorithm (FDTD algorithm) is used to optimize the cell structure. Introducing the metasurface into the spectral imaging systems, and the research on the metasurface spectral imaging system is carried out by optimizing the size, and structure arrangement of the sub-wavelength structure to realize the independent regulation of multi-wavelength dispersion and to focus. Using this method, scanning the influence of different structure on the phase,according to the phase distribution of the hyperbolic plane lens, several different focusing hyperlenses are designed for different wavelengths, achieving ametasurface multispectral imaging system with eight spectral segments in the visible band 510~710 nm. Electromagnetic and optical simulation software(FDTD solutions), the data processing software is used to analyze the far-field electric field intensity data to obtain spectral data of different wavelengths. Metasurfaces provide a new way for spectral imaging technology and have great application potential in miniaturized spectral remote sensing fields such as aerospace. Compared with traditional grating or prism spectroscopy structures, metasurface spectacle imaging systems can effectively reduce the system’s volume. Their ultra-light, ultra-thin, and portable characteristics solve the limitations of existing spectroscopy imaging systems and provide a theoretical basis for developing miniaturized and lightweight spectroscopy systems.
		2023 Vol. 43 (01): 225-229 [Abstract] ( 242 ) RICH HTML PDF (2397 KB) ( 143 )

	230	Study on Reflection Characteristics of Completely Oxidized DZ125
		YE Zhi-peng^{1, 2, 3}, ZHAO Shu-nan⁴, LI Xun-feng^{1, 2, 3*}, HUAI Xiu-lan^{1, 2, 3}
		DOI: 10.3964/j.issn.1000-0593(2023)01-0230-09
		As the core component of a gas turbine, the turbine blade works at a high temperature of thousands of degrees for a long time. In order to ensure the safe and reliable bladeoperation, its temperature needs to be monitored in real-time. Radiation temperature measurement is currently the mainstream method of non-contact temperature measurement of gas turbine blades. Its temperature measurement accuracy is closely related to the reflection characteristics of blade materials. The difficulty of current research isto predict the reflected energy in different directions and reduce the impact of reflected radiation on temperature measurement. In order to predict the reflected energy of the turbine blade and improve the accuracy of radiation temperature measurement, the bidirectional reflection distribution function (BRDF) of completely oxidized DZ125, a common material of turbine blade, is studied in this paper.The reference method is used as the experimental measurement method. Firstly, the measurement principle and data processing method of the BRDF comparison method is analyzed. After that, the experiment platform was built independently. Under the conditions of 25, 900 and 1 100 ℃ and wavelength of 1 060, 1 550 and 1 908 nm, the incident zenith angle and reflection zenith angle were controlled to change in the range of 0° to 60°, and the azimuth angle changed in the range of 0° to 180°. Several groups of the BRDF values were measured and calculated, and the effects of various factors on BRDF of completely oxidized DZ125 were analyzed. Finally, the modified Phong model is used to fit the measured values of BRDF. The fitting results can be accepted accordingto the comparison with the experimental results. The results show that temperature and wavelength have little effect on BRDF of fully oxidized DZ125. As long as the temperature range of the turbine blade and the radiation wavelength does not change strongly, the effect of both on BRDF can be ignored.The incident zenith angle, reflection zenith angle and azimuth angle greatly affect BRDF. The closer the incident zenith angle and reflection zenith angle are,and the closer the azimuth angle is 180°, the closer the position of the measuring probe is to the specular reflection direction. Then the specular reflection characteristics of the sample are more significant, and the BRDF is large. On the contrary, diffuse reflection dominates when the probe is far away from the specular reflection direction, and BRDF decays rapidly. The experimental results show that completely oxidized DZ125 has strong specular reflection characteristics. The modified Phong model fitting results show the bidirectional reflection characteristics of fully oxidized DZ125. This model is simple and suitable for using Monte Carlo simulation to predict the distribution of the total energy of DZ125 oxidation.It provides a theoretical basis for the subsequent study of the radiation temperature measurement of turbine blades.
		2023 Vol. 43 (01): 230-238 [Abstract] ( 91 ) RICH HTML PDF (8641 KB) ( 32 )

	239	Research on Quantitative Regression Method of IR Spectra of Organic Compounds Based on Ensemble Learning With Wavelength Selection
		JU Wei¹, LU Chang-hua^{2, 3}, ZHANG Yu-jun³, CHEN Xiao-jing¹, JIANG Wei-wei^2*
		DOI: 10.3964/j.issn.1000-0593(2023)01-0239-09
		The application of the ensemble learning method in the quantitative analysis of organic infrared spectra and the influence of the characteristic wavelength selection method on the modeling efficiency and prediction accuracy of infrared spectra ensemble learning is studied. Taking the cetane number and total aromatic hydrocarbon content of diesel infrared spectra as the research object, firstly, a two-layer stacking ensemble learning framework is established by using extreme random forest (ERT), linear kernel support vector machine (LinearSVM), radial basis kernel support vector machine (RBFSVM) and polynomial kernel support vector machine (polySVM) as baselearners, and LinearSVM as meta-learners. The quantitative regression accuracy of diesel infrared spectra by single base learners and ensemble learning model is analyzed and compared. Compared with the partial least squares (PLS) quantitative regression model, the prediction accuracy of the Stacking ensemble learning model for two organic compounds in diesel spectra is improved. The ERT model for cetane number content is the best (r=0.848, RMSEP=1.603, RDP=2.627), the prediction result of Stacking model for total aromatic content is the best (r=0.991, RMSEP=0.645, RDP=9.243). Further, the characteristic wavelengths of infrared spectra are selected using the combined partial least squares (SiPLS) and successive projections algorithm (SPA), and the ensemble learning quantitative regression model is established using the selected characteristic wavelengths. Among them, the prediction results of the SiPLS-ERT model for cetane number content are the best (r=0.893, RMSEP=1.013, RDP=3.051), and the prediction results of the SiPLS-Stacking model for total aromatic content are the best (r=0.998, RMSEP=0.354, RDP=11.475), and the average training time of the model is reduced by more than 50% compared with the full spectra training time, and the modeling speed is significantly improved. The results show that the characteristic wavelengths combined with ensemble learning quantitative regression modeling can be used in the quantitative analysis of organic infrared spectra. Compared with the traditional quantitative regression method, the modeling efficiency and prediction accuracy of this method are greatly improved, which provides relevant method support for the further study of the application of machine learning in the quantitative analysis of spectra.
		2023 Vol. 43 (01): 239-247 [Abstract] ( 106 ) RICH HTML PDF (6526 KB) ( 80 )

	248	Green Preparation of Biomass Carbon Quantum Dots for Detection of Cu²⁺
		LIU Yu-ying^{1, 2}, WANG Xi-yuan^{1, 2*}, MEI Ao-xue^{1, 2}
		DOI: 10.3964/j.issn.1000-0593(2023)01-0248-06
		Sunflower is one of the main oil crops in China. Its straw is a natural cellulose material, which has the advantages of being green, non-toxic and low cost, and it is an ideal material for synthesising biomass carbon quantum dots. In recent years, due to the abuse of copper-containing pesticides and chemical fertilizers, many copper-containing pollutants have been discharged, resulting in the copper content in farmland soil and water environment being much higher than the environmental background value. Therefore, it is urgent to develop a Cu²⁺ detection method with good selectivity, high sensitivity and environmental friendly. Carbon quantum dots (CDs) are quasi-spherical fluorescent carbon nanomaterials with particle sizes less than 10 nm. Its surface contains abundant polar functional groups and has been widely studied for its good water solubility. Compared with traditional semiconductor quantum dots (CdSe, CdTe), CDs have extensive synthetic materials and good biocompatibility advantages. They are mainly used in biological imaging, photo catalysis, photoelectric conversion, sensor detection and other fields. However, most of the precursors of carbonized CDs synthesis are expensive chemicals, which have the disadvantages of a complex synthesis processes and environmental pollution, limiting the large-scale production and application of CDs. Therefore, developing an eco-friendly, simple and inexpensive CDs synthesis method is of great significance. In this study, waste sunflower straw was used as a carbon source, and a simple hydrothermal method was used to synthesize biomass carbon quantum dots (S-CDs) as fluorescent probes to detect Cu²⁺. Through a series of optical properties analysis and characterization of S-CDs, the surface functional groups of S-CDs mainly include O—H, N—H andC═N, among which abundant O—H can provide colloid stability, effectively control the morphology and particle size distribution of nanoparticles, thus improving the quantum yield (QYs) of S-CDs. ex=317 nm, em=456 nm, S-CDs have excellent optical performance and good optical stability in the pH range of 2.0~12.0, less affected by high salinity environment, QYs is about 8.42%, emitting blue fluorescence under UV analyzer irradiation at 365 nm. In addition, the fluorescence quenching effect of S-CDs induced by Cu²⁺ was further studied using the synthesized S-CDs as a fluorescence probes. The results showed that the prepared S-CDs were sensitive to Cu²⁺ in the concentration range of 0～10 μmol·L^-1 with a good linear relationship (R²=0.971 4), and the detection limit (LOD) was as low as 167 nmol·L^-1. In practice, the detection recoveries of Cu²⁺ in lake water are 96.18%～109.22%. In this study, a Cu²⁺ detection method based on fluorescent carbon quantum dots was introduced based on resource utilization of straw waste.
		2023 Vol. 43 (01): 248-253 [Abstract] ( 141 ) RICH HTML PDF (4208 KB) ( 75 )

	254	Research on Testing NH₃-N and COD in Water Quality Based on Continuous Spectroscopy Method
		LI Wen, CHEN Yin-yin^*, LUO Xue-ke, HE Na
		DOI: 10.3964/j.issn.1000-0593(2023)01-0254-06
		Aiming at the requirements for accurate and rapid combined determination of ammonia nitrogen and chemical oxygen demand in surface waters of class Ⅰ—Ⅴ, groundwater and industrial wastewater, this paper combines continuous spectrometry and Sequential Injection Analysis (SIA) in spectral Analysis, based on the national standard for surface water detection, taking in-situ water quality parameters of ammonia nitrogen (NH₃-N) and chemical oxygen demand (COD) as the detection objects, designing a micro, efficient and rapid detection instrument for NH₃-N and COD in-situ water. The system mainly relies on the self-designed digestion cell structure based on ultraviolet lamp digestion with heated closed digestion, and the detection cell structure based on the spectral scanning design to achieve the purpose of rapid digestion and stable detection. It also optimized the detection process based on spectrophotometry. At the beginning of COD digestion, the coordination compounds in the detection tank after NH₃-N index coloritization were determined by spectral scanning. After digestion, COD was determined, the whole detection process was shortened by at least 60 minutes compared with the national standard detection method. It can automatically complete the determination of NH₃-N and COD within 25 minute, greatly saving time cost. Plotting the absorbance and continuous wavelength curve of coordination compounds after spectral scanning color reaction: NH₃-N and COD have obvious absorption peaks at 690 and 445 nm, respectively. After reading the absorbance value at the peak, the least square method is used to establish regression modeling for NH₃-N and COD, fitting the regression equation and calculating the correlation coefficient, and drawing the absorbance-concentration working curve of the corresponding parameters. The experimental results show that the correlation coefficient r of the NH₃-N standard working curve is≥0.998 7 in the concentration range of 0~2 mg·L^-1, and the concentration is positively correlated with the absorbance. The relative standard deviations of repeatability were 1.36%～1.68%, and the recoveries were 97%~102%. In the range of 0~50 mg·L^-1, the correlation coefficient r of COD standard working curve is ≥0.997 8, and the concentration is negatively correlated with the absorbance. The relative standard deviations of repeatability were 2.14%~2.48%, and the recoveries were 97.6%~102.95%. The system is accurate, linear and stable, and has high feasibility and reliability. Research on the method of combined determination of NH₃-N and COD based on SIA and continuous spectroscopy is of great value in the research to broaden the application of spectroscopy in the field of rapid water quality testing and to improve the efficiency of detection.
		2023 Vol. 43 (01): 254-259 [Abstract] ( 88 ) RICH HTML PDF (2756 KB) ( 43 )

	260	Research on ICP-AES Spectral Baseline Correction Method Based on DE Algorithm and NURBS Curve
		LIAN Xiao-qin^{1, 2}, CHEN Yan-ming^{1, 2}, WANG Yu-qiao^{1, 2}, LIU Yu^{1, 2}
		DOI: 10.3964/j.issn.1000-0593(2023)01-0260-08
		Inductively Coupled Plasma Atomic Emission Spectroscopy (ICP-AES) analysis method is a commonly used solution element concentration analysis method. However, in the process of ICP-AES measurement, due to temperature drift, stray light and instrument dark current, etc., the spectrum often has a certain degree of the baseline drift phenomenon, resulting in errors in the measurement results of element spectral intensity values, which in turn affects the quantification of element concentrations. Therefore, baseline correction is one of the necessary links in the ICP-AES analysis method. In this paper, the traditional spectral baseline correction method is briefly analyzed, and on this basis, a ICP-AES spectral baseline drift correction method based on non-uniform B-spline curve and differential evolution algorithm is designed; Firstly, it is verified that the probability density distribution of noise in the spectral signal obeys the Gaussian distribution, then the original spectrum is preprocessed, and the spectral signal is denoised by Gaussian filtering; Then, the standard deviation of the minimum value sequence in the process of spectral baseline correction is used as the evaluation index, the non-uniform B-spline curve is used as the baseline model, and the control point sequence C and the internal point sequence T of the curve are used as the characteristic parameters of the evaluation function to establish the ICP- AES spectral baseline correction evaluation function, so that the spectral baseline correction problem can be converted into the problem of solving the global optimal solution of the characteristic parameters of the evaluation function; Finally, this paper briefly introduces the process of the differential evolution algorithm, and uses the differential evolution algorithm to solve the global optimal solution of the characteristic parameters when the evaluation function achieves the minimum value, that is, the control point sequence C and the interior point sequence T of the non-uniform B-spline curve, and fit the corresponding non-uniform B-spline curve as the spectral baseline to realize the baseline correction of the ICP-AES spectrum. A dataset of measured ICP-AES spectral data is used to verify the baseline drift correction method proposed in this paper. The experimental results show that the ICP-AES spectral baseline correction method was based on the differential evolution algorithm. And the non-uniform B-spline curve proposed in this paper can accurately calculate the control point sequence C and the internal point sequence T of the non-uniform B-spline curve and fit the suitable spectral baseline, which enables baseline correction of ICP-AES spectra. The method can overcome the limitation of the non-uniform B-spline curve in the field of spectral baseline correction and provide a technical basis for the subsequent quantitative analysis of element content.
		2023 Vol. 43 (01): 260-267 [Abstract] ( 93 ) RICH HTML PDF (5081 KB) ( 34 )

	268	Clustering Analysis of FTIR Spectra Using Fuzzy K-Harmonic-Kohonen Clustering Network
		CHEN Yong^{1, 2}, GUO Yun-zhu¹, WANG Wei^3, WU Xiao-hong^{1, 2}, JIA Hong-wen⁴, WU Bin⁴
		DOI: 10.3964/j.issn.1000-0593(2023)01-0268-05
		Different foods contain different nutrients and effectiveness, and there are differences in their Fourier transform infrared spectra. In order to classify varieties of foods correctly, this paper presented the way to classify varieties by combining Fourier transform infrared spectroscopy (FTIR) with fuzzy clustering analysis. Fuzzy K-harmonic Kohonen clustering network (FKHKCN) was proposed by introducing fuzzy K-harmonic means (FKHM) clustering into the learning rate and update strategy of the Kohonen clustering network. The learning rate of FKHKCN is computed by fuzzy membership values of fuzzy C-means (FCM) clustering, and the cluster centers of FKHKCN can be derived from the cluster centers of FKHM. Therefore, FKHKCN can solve the problem that the Fuzzy Kohonen clustering network (FKCN) is sensitive to the initial cluster centers, and the clustering result is unstable. FKHKCN can achieve the clustering analysis of FTIR data as a fuzzy clustering algorithm. This experiment involves three datasets: (1) Three kinds of tea samples (Emeishan Maofeng, good and poor Leshan trimeresurus) were obtained from Sichuan, China as experimental samples with a total number of 96. (2) Two kinds of coffee samples (robusta and arabica). (3)Three meat samples (chicken, pork and turkey). To start with, three datasets were preprocessed. Scattering effects in the original spectra data of tea samples were reduced by multiple scattering correction (MSC). Savitzky-Golay was used to reduce noise in FTIR spectra of coffee and meat samples. Secondly, the high dimensional FTIR data of three datasets were reduced to by the low dimensionaldata by principal component analysis (PCA). Thirdly, tea data after PCA were extracted by linear discriminant analysis (LDA) and the spectral data were projected into the obtained discriminant vectors. Finally, FCM, FKCN and FKHKCN were used to classify the three datasets, respectively. The experimental results showed that FCM, FKCN and FKHKCN achieved the clustering accuracies for the tea varieties with the values: 90.91%, 90.91% and 93.94%, respectively; the clustering accuracies for the meat varieties with the values: 90.83%, 0.00% and 92.50%, respectively; the clustering accuracies for the coffee varieties with the values: 89.17%, 89.17% and 90.83%, respectively. The above experimental results indicated that FTIR technology coupled with PCA, LDA and FKHKCN was an effective method for classifying food varieties, and its clustering accuracy was higher than FCM and FKCN, and its clustering result was stable.
		2023 Vol. 43 (01): 268-272 [Abstract] ( 82 ) RICH HTML PDF (1006 KB) ( 42 )

	273	Tillering Number Estimation of Winter Wheat Based on Visible Spectrogram and Lightweight Convolutional Neural Network
		LI Yun-xia¹, MA Jun-cheng², LIU Hong-jie³, ZHANG Ling-xian^1*
		DOI: 10.3964/j.issn.1000-0593(2023)01-0273-07
		Tiller number is a key trait to characterize the growth of winter wheat, which is of great significance for seedling condition monitoring and yield prediction of winter wheat. Given the complicated data acquisition and thelarge volume of the estimation model, an estimation method for the tiller number of winter wheat based on visible light images and lightweight convolutional neural networks (CNNs) was explored. It can realize nondestructive and rapid estimation of tillering numbers and can be embedded into mobile terminal devices. It can realize nondestructive and rapid estimation of tillering number of winter wheat and can be embedded into mobile terminal devices. Based on these data, lightweight CNNs MobileNetV2, SqueezeNet and ShuffleNet were used to construct the estimation model of tillering number of winter wheat. The optimization comparison test was conducted, and the comparison test was conducted with the estimation model constructed based on non-lightweight neural network AlexNet and ResNet series models. Comparison tests were conducted with estimation models constructed based on non-lightweight CNNs, AlexNet and ResNet series models. In addition, the robustness of the tillering number estimation model under different plant densities and the generalization ability of data in different growing seasons were verified. The results showed that the determination coefficient (R²) and normalized root mean square error (NRMSE) of the estimation model based on MobileNetV2 were 0.7 and 0.2, respectively, which showed the best performance among the three lightweight CNNs. The volume of the winter wheat tillering number estimation model constructed based on non-lightweight CNNs was 2.3~16.1 times that of the winter wheat tillering number estimation model constructed based on MobileNetV2. Compared with the non-lightweight CNNs, the estimation model based on MobileNetV2 had better R² and smaller volume, which was suitable for embedding into mobile terminal devices. According to the visible image data set divided by three plant densities, 120, 270 and 420 plants ·m^-2, the value of R² of the estimation model based on MobileNetV2 were 0.8, 0.8 and 0.7, respectively, showing robust performance. For visible images of two growing seasons, the estimation model based on MobileNetV2 improved R² by 2 times, and NRMSE decreased by 7.6% through transfer learning, showing good adaptability to seasonal differences of data and reflecting the generalization ability of the model. Therefore, based on visible light images, the lightweight CNNs estimation model can meet the tillering number estimation of winter wheat and provide an accurate, robust tool that can be embedded into mobile terminal devices for winter wheat growth observation and field agronomic measures management decisions.
		2023 Vol. 43 (01): 273-279 [Abstract] ( 93 ) RICH HTML PDF (3087 KB) ( 45 )

	280	Detection of Dairy Cow Mastitis From Thermal Images by Data Enhancement and Improved ResNet34
		ZHANG Qian¹, YANG Ying^1*, LIU Gang^{1, 2, 3}, WU Xiao¹, NING Yuan-lin¹
		DOI: 10.3964/j.issn.1000-0593(2023)01-0280-09
		Mastitis is one of the most serious diseases in dairy production and breeding. The early detection of cow mastitis can provide the basis for follow-up treatment to improve the efficiency of disease treatment and reduce the risk of breeding. In order to realize fast and high-precision “one-step” mastitis disease detection for naturally walking dairy cows, a dairy cow mastitis disease detection method based on the thermal infrared image, data enhancement and improved ResNet34 is proposed in this paper. Compared with the existing “multi-step” dairy cow infrared image mastitis detection method, this method does not need the positioning of key parts of dairy cows, such as breast and eyes and temperature extraction, which can effectively avoid the error accumulation caused by “multi-step”, to achieve more efficient mastitis detection. Firstly, this method horizontally splices the local pictures containing the key parts of the cow into an overall picture with complete information and expands the training samples combined with the RandAugment data enhancement method; Secondly, the ResNet34 residual network is used as the basic network of the experiment, and the model is improved as follows according to the characteristics of thermal infrared image: (1) simplify the redundant internal layer of the network to make the model lighter; (2) Auxiliary classifiersare added in the middle layer to make up for the feature loss caused by model simplification; (3) The improved multi fusion pool layer is used to replace the original single pool layer, which makes the content of feature extraction richer. Finally, 3 298 thermal infrared images (66 cows) were randomly selected as the experimental objects, and multiple groups of comparative experiments were set. The results showed that compared with the traditional ResNet34, the classification accuracy of the improved ResNet34 model was improved by 3.4%, the model verification accuracy based on the improved ResNet34 combined with transfer learning and data enhancement was 90.3%, the test accuracy was 88.4%, and the classification time was only 3.39×10^-3 seconds. In addition, to ensure theindependence of the experimental data set, this paper further divides it into the training set, verification set and test set according to the number of dairy cows in 3∶1∶1. The test accuracy of the model was 80.3%, which proves that the proposed model has good robustness. According to the test results, it is calculated that the precision rate, recall rate and F1 score of the model are 91.2%, 91.6% and 91.4%. Compared with previous experiments, the accuracy is improved by 5.1% and the specificity is improved by 5.3%. In conclusion, this research method can provide a reference for screening and medical diagnosis of breast diseases in early dairy cows.
		2023 Vol. 43 (01): 280-288 [Abstract] ( 135 ) RICH HTML PDF (4504 KB) ( 47 )

	289	A Multi-Task Convolutional Neural Network for Infrared and Visible Multi-Resolution Image Fusion
		ZHU Wen-qing^{1, 2, 3}, ZHANG Ning^{1, 2, 3}, LI Zheng^{1, 2, 3*}, LIU Peng^{1, 3}, TANG Xin-yi^{1, 3}
		DOI: 10.3964/j.issn.1000-0593(2023)01-0289-08
		Infrared and visible image fusion have always been a research hotspot in the image field. Fusion technology can compensate for a single sensor’s deficiency and provide good imaging pandation for image understanding and analysis. Due to the limitation of production technology and cost, the resolution of infrared detectors is much lower than that of visible detectors, which prevents practical usage to a great extent. A multi-task convolutional neural network framework combining infrared super-resolution and image fusion tasks is proposed, which is applied to the infrared and visible multi-resolution image fusion. In terms of network structure, firstly, a dual-channel network is designed to extract infrared and visible features respectively, so that the resolution of each source image does not limit the proposed algorithm. Secondly, the feature up-sampling block is proposed, using the bilinear interpolation method to increase the number of pixels. Then the mapping relationship between pixel smooth space and high-frequency space is refined via a multilayer perceptron. Therefore, the infrared images can be presented on an arbitrary scale, where the training tasks are not provided. Furthermore, the linear self-attention mechanism is introduced into the network to learn the nonlinear relationship between feature space positions, suppress irrelevant information and enhance global information expression. In terms of the loss function, the gradient loss is proposed to retain the filter response with larger absolute values in the infrared and visible images and calculate the Frobenius norm between the value and the response value of the reconstructed fusion image. Thus, fusion images can be generated without ideal images as ground truth supervising network learning. Finally, the fused and high-resolution infrared images can be reconstructed simultaneously by optimizing the multi-task model under the combined action of gradient loss and pixel loss. The proposed approach is trained on the RoadScene dataset and compared with the other four related algorithms on the TNO dataset. In terms of subjective performance, the proposed method can input source images with the arbitrary resolution, and fusion images have prominent infrared targets and rich visible details. When the resolution of source images is quite different, the proposed method can still reconstruct high-resolution infrared images with clear features and has robust generalization. The objective performance is excellent in multiple evaluation metrics such as entropy, the sum of the correlations of differences and spatial frequency. Experimental results demonstrate that fusion images have a large amount of information, high information conversion rate and high clarity, which verifies the effectiveness of the proposed method.
		2023 Vol. 43 (01): 289-296 [Abstract] ( 121 ) RICH HTML PDF (6353 KB) ( 117 )

	297	Research on Hyperspectral Features and Recognition Methods of Typical Camouflage Materials
		HU Yi-bin¹, BAO Ni-sha^{1, 2*}, LIU Shan-jun^{1, 2}, MAO Ya-chun^{1, 2}, SONG Liang³
		DOI: 10.3964/j.issn.1000-0593(2023)01-0297-06
		Aiming at the phenomenon of “foreign objects with the same spectrum” in the camouflaged target and the background target in certain specific environments, traditional visible light and multi-spectral remote sensing technologies have limitations in camouflage recognition. This paper, applies hyperspectral technology to the characteristic analysis and recognition of typical camouflage materials. The SVC HR1024 spectrometer was used to obtain the Visible-NIR Spectrum of the jungle camouflage net under different water immersion times. The spectral characteristics and sensitive bands of the jungle camouflage net under different water immersion conditions and typical vegetation in northern China were analyzed and revealed through spectral similarity measurement and envelope removal treatment. Based on the near-infrared band, the spectral ratio index (RCI) was constructed to identify the camouflaged targets in the green vegetation environment. Finally, the hyperspectral image in the simulation camouflage environment was obtained through a hyperspectral imaging experiment, and the recognition effect was verified using the hyperspectral image. The results showed that: (1)The basic morphology of the spectral curve of the jungle camouflage net with different water immersion times was similar, and its reflectivity decreased as a whole with increasing water immersion time. The 1 900 nm band is the most obvious band that the reflectance spectrum of jungle camouflage net responds to water content, and its spectral characteristics are similar to those of vegetation due to water immersion treatment, and the similarity is increased from 0.895 to 0.939. (2)The similarity between camouflage net and vegetation is high in the visible band, and the spectral fluctuation is similar, but the spectral characteristics of the jungle camouflage net and vegetation are different in near-infrared band. Through the analysis of the envelope removal method, it is concluded that the bands around 970, 1 190 and 1 440 nm are sensitive bands for identifying the jungle camouflage net. Moreover, based on the two obvious differences in reflectance slope between the jungle camouflage net and the vegetation in the band range of 900~1 900 nm, RCI₁ (R_{1 190}/R_{1 270}) and RCI₂ (R_{1 270}/R_{1 440}) were constructed. (3) The decision tree classification model based on the RCI index can quickly and effectively extract the camouflaged target from the green vegetation background. Experimental results show that using the RCI index to identify and extract the camouflaged target area, the results obtained are in good agreement with the original image in shape and size, and the recognition accuracy can reach 95%, indicating that the index has a good recognition effect on camouflaged targets.
		2023 Vol. 43 (01): 297-302 [Abstract] ( 141 ) RICH HTML PDF (3655 KB) ( 96 )

	303	Rapid Qualitative Analysis of Wool Content Based on Improved U-Net++ and Near-Infrared Spectroscopy
		LENG Si-yu^{1, 2}, QIAO Jia-hui¹, WANG Lian-qing³, WANG Jun^{1, 2*}, ZOU Liang¹
		DOI: 10.3964/j.issn.1000-0593(2023)01-0303-07
		Wool products are popular because of their softness and warmth. The content of wool is an important indicator of the quality of wool products. However, the quality of wool products in the market varies. In addition, traditional testing methods are destructive, and the results might be subjective, which can no longer meet the need to evaluate the quality of the target wool products quickly. NIR spectroscopy is a rapid measurement method that does not require the destruction of sample structure and can be embedded with machine learning models. Because of this, this paper proposes a rapid qualitative wool content evaluation method via fusing NIR spectroscopy and attention-based U-Net++. In terms of data preparation, this paper employs a handheld portable spectrometer to collect spectral data of wool product samples with a wavelength range of 908.1 to 1 676.2 nm. The original samples are graded according to their contents. The experiments collected spectral datasets of the same sample at 5 heights of 5, 6, 8, 9 and 19 mm from the spectrometer, and abnormal samples were removed by Mahalanobis distance. 5 125 sets of spectral data were used for the final data modeling. Regarding model selection, the U-Net++ network provides an end-to-end way for feature extraction and classification with down-sampling, jump connections and up-sampling operations. However, due to alarge number of skip connections, it reuses low-level features, and the models might contain redundant parameters. This paper introduces an attention-gating module which can extract feature information more effectively and improve prediction accuracy. The spectral data corresponding to 90% of wool product samples is used for training and validation, and the rest spectral data is used for testing. The experimental results show that the prediction model based on the U-Net++ network obtains an accuracy of 93.59%, a recall of 93.53%, and a precision of 94.24% on the independent test set, all of which outperform traditional classification models. Meanwhile, the classification model proposed in this paper outperforms other U-Net series networks, such as U-Net and Attention U-Net, demonstrating the effectiveness of the skip connection and attention-gating modules. In this paper, the spectral analysis based on the Attention U-Net++ model and portable near-infrared spectrometer provides a practical and meaningful way for rapid, nondestructive inspection of wool content.
		2023 Vol. 43 (01): 303-309 [Abstract] ( 114 ) RICH HTML PDF (2027 KB) ( 81 )

	310	Deep Learning Modelling and Model Transfer for Near-Infrared Spectroscopy Quantitative Analysis
		FU Peng-you^{1, 2}, WEN Yue², ZHANG Yu-ke³, LI Ling-qiao^1, YANG Hui-hua^{1, 2}
		DOI: 10.3964/j.issn.1000-0593(2023)01-0310-10
		Near-infrared spectroscoqy analysis technologyrelies on Chemometric methods that characterize the relationships between the spectral matrix and the chemical or physical properties. However, the samples’ spectra are composed of signals and various noises. It is difficult for traditional Chemometric methods to extract the effective features of the spectra and establish a calibration model with strong generative performance for a complex assay. Furthermore, the same quantitative analysis results cannot be achieved when the calibration model established on one instrument is applied to another because of the differences between the instruments. Hence, this paper presents a quantitative analysis modeling and model transfer frameworkbased on convolution neural networks and transfer learning to improve model prediction performance on one instrument and across the instrument. An advanced model named MSRCNN is presented based on a convolutional neural network, which integrates multi-scale feature fusion and residual structure and shows outstanding model generalization performance on the master instrument. Then, four transfer learning methods based on fine-tuning are proposed to transfer the MSRCNN established on the master instrument to the slave instrument. The experimental results on open accessed datasets of drug and wheat show that the RMSE and R² of MSRCNN on the master instrument are 2.587, 0.981, and 0.309, 0.977, respectively, which outperforms PLS, SVM, and CNN. Byusing 30 slave instrument samples, the transfer of the convolutional layer and fully connected layer in the MSRCNN model is the most effective among the four fine-tune methods, with RMSE and R² 2.289, 0.982, and 0.379, 0.965, respectively. The performance can be further improved by increasing the sample of slave instruments that participated in model transferring.
		2023 Vol. 43 (01): 310-319 [Abstract] ( 195 ) RICH HTML PDF (4370 KB) ( 181 )

	320	Improved Sensitivity of Localized Surface Plasmon Resonance Using Silver Nanoparticles for Indirect Glyphosate Detection Based on Ninhydrin Reaction
		XU Meng-lei^{1, 2}, GAO Yu³, ZHU Lin¹, HAN Xiao-xia¹, ZHAO Bing^1*
		DOI: 10.3964/j.issn.1000-0593(2023)01-0320-04
		Conventional pesticide residue detection still suffers from numerous steps, is long time-consuming, and insufficient delicacy. According to ninhydrin colouring and the principle of localized surface plasmon resonance (LSPR) enhancing absorption, glyphosate in water samples can be detected by ultraviolet-visible spectroscopy (UV-Vis). Furthermore, density functional theory is used to analyze the enhancement mechanism of absorption of purple color dye (PD) products. The PD product displays a maximum absorption of around 570 nm when glyphosate reacts with ninhydrin is detected by UV-Vis. There are slight shifts from 570 nm in the UV-Vis spectrum to 568 nm with a stronger peak when the PD product is absorbed on Ag NPs, and the limit of detection at 2.017 4×10^-11 mol·L^-1, which is much lower than 6.5×10^-7 mol·L^-1 limit of detection reported. Gaussian 09 software carried out that the PD product would attach to Ag NPs via an Ag—O bond through ninhydrin’s C═O group vertically. MEP mapping provides C═O group interaction with Ag NPs stable, C═O group coupled C—N comprise a large π-conjugated system in their plane. The color of the C—N group is blue, which suggests that C—N coupled with the C═O group are the chromophore in the PD product. Thus, an indirect detection method derived from ninhydrin can be used for glyphosate detection in water samples. The LSPR effect of Ag NPs enhances the absorption intensity with higher sensitivity than the conventional method.
		2023 Vol. 43 (01): 320-323 [Abstract] ( 98 ) RICH HTML PDF (1382 KB) ( 62 )

	324	Antiglycation Activity on LDL of Clove Essential Oil and the Interaction of Its Most Abundant Component—Eugenol With Bovine Serum Albumin
		LI Jin-zhi¹, LIU Chang-jin^{1, 4}, SHE Zhi-yu², ZHOU Biao², XIE Zhi-yong², ZHANG Jun-bing³, JIANG Shen-hua^{2, 4}
		DOI: 10.3964/j.issn.1000-0593(2023)01-0324-09
		The previous research results of our laboratory showed that clove had the strongest antioxidant activity in the state’s list of food-medicine herbs promulgated. It has been found that the antioxidant activity of natural products is closely related to antiglycation activity. Therefore, this study aimed to investigate further the antiglycation activity of clove essential oil (CEO) and the interaction between eugenol the component with the highest content in CEO and bovine serum albumin (BSA). The spectral results showed that in the non-enzymatic glycation incubation system of low-density lipoprotein (LDL), CEO had significant inhibition effects on forming the early, intermediate and late products of LDL glycation and had the strongest inhibition effect on the late product. The composition analysis of CEO by gas chromatography-mass spectrometry (GC-MS) indicated that eugenol was the most abundant component in CEO. Multispectral and molecular docking were applied to investigate the interaction between eugenol and BSA. The result of ultraviolet-visible (UV-Vis) absorption spectroscopy indicated an interaction between eugenol and BSA. In fluorescence emission spectroscopy, with the increase of eugenol concentration, the fluorescence intensity of BSA gradually increased with the blue shift, which further proved the interaction between them. The calculated results of binding parameters at different temperatures confirmed that the thermodynamic processes were involved in the interaction between eugenol and BSA. Thermodynamic parameters and the site marker competitive experiments showed that eugenol particularly bonds to BSA at the site Ⅰ through hydrogen bond and vander Waals force. In synchronous fluorescence (SF), three-dimensional (3D) fluorescence, and Fourier transform infrared (FTIR) spectroscopy, the signal strengths changed with the increase of eugenol concentration, and the shifts also occurred, which indicated that the conformation of BSA changed with the addition of eugenol. The results of the interaction between eugenol and BSA were further verified by molecular docking technology. The findings can provide theoretical support for the further development of clove.
		2023 Vol. 43 (01): 324-332 [Abstract] ( 115 ) RICH HTML PDF (4238 KB) ( 49 )

　CopyRight © OFFICE OF SPECTROSCOPY AND SPECTRAL ANALYSIS
　Address: NO.76 Collage South Road of Beijing 　Postalcode:100081
　Tel. (010)62181070　E-mail:chngpxygpfx@vip.sina.com
　Support by Beijing Magtech Co.ltd　E-mail:support@magtech.com.cn