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

2025 Vol. 45, No. 07 Published: 2025-07-01

	1801	Coherent Raman Scattering Microscopy and Its Recent Research Progress in in Vivo Imaging
		LI Shu-qi, LUO Guo-quan, CHEN Yu, YU Bin, QU Jun-le, LIN Dan-ying^*
		DOI: 10.3964/j.issn.1000-0593(2025)07-1801-08
		In vivo imaging technology has revolutionary significance, enabling real-time observation of living organisms and providing dynamic information directly related to biological processes. This is crucial for understanding disease mechanisms and evaluating treatment effects. Coherent Raman scattering (CRS) microscopy offers the advantage of specific molecular imaging in biological samples without the need for fluorescent probes that might interfere with biomolecular function, making it a promising tool in the field of in vivo imaging. However, CRS microscopy still faces challenges in practical applications in in vivo imaging, including photodamage, limited imaging depth, and motion artifacts. Recent advancements in related technology have led to significant breakthroughs, addressing these challenges by minimizing photodamage, extending imaging depth, eliminating or reducing motion artifacts, and enabling multimodal imaging. In vivo real-time imaging of human skin, brain, and spine in experimental animal models and tumors has driven substantial progress in CRS microscopy, both in in vivo imaging research and clinical applications. This paper offers a comprehensive review of the latest developments in CRS microscopy for in vivo imaging, providing an in-depth analysis of current challenges and their solutions to contribute to this technology's ongoing development and broader application. Common strategies to overcome photodamage involve reducing the thermal effects and chemical reactions induced by the laser in the sample, typically by limiting laser power and integration time. Several approaches have been explored to address the limitation of imaging depth, including imaging superficial tissues such as the skin or areas near the surface, combining optical windows, or directly imaging deeper tissues or organs exposed through minimally invasive surgery. Adaptive optics technology helps balance depth with non-invasive imaging, while endoscopic imaging provides an additional solution. To minimize or eliminate motion artifacts, it is crucial first to keep the organism stationary or reduce movement through appropriate anesthesia and fixation techniques. In addition, optical windows and real-time motion correction algorithms can be employed to mitigate further jitter caused by physiological activities like breathing and heartbeat in anesthetized samples. Increasing imaging speed is another way to reduce motion artifacts. Finally, combining CRS with other nonlinear optical microscopy techniques, such as two-photon excitation fluorescence and second-harmonic generation, enables multimodal imaging, providing richer information, enhancing the analysis of in vivo biological samples, and offering deeper insights into biological processes.
		2025 Vol. 45 (07): 1801-1808 [Abstract] ( 8 ) PDF (14487 KB) ( 3 )

	1809	Research on Spectral Reconstruction Based on Camera Response Prediction
		LIANG Jin-xing^{1, 2}, ZHOU Wen-sen¹, HU Xin¹, LI Yi-fan¹, WANG Ding-kang¹, LI Ning¹, PENG Tao^{1, 2*}
		DOI: 10.3964/j.issn.1000-0593(2025)07-1809-11
		The surface spectral reflectance of the object is regarded as the fingerprint of its color. At the same time, it is also important to characterize the physical and chemical properties of substances. Spectral imaging technology based on spectral reconstruction can overcome the dependence of RGB images on imaging conditions. Meanwhile, it can effectively improve the spectral image's spatial resolution and acquisition efficiency and reduce equipment costs. Different from the principle of multispectral cameras, spectral imaging based on spectral reconstruction first captures the digital images of the object using a digital imaging system, and then the corresponding spectral images are reconstructed using spectral reconstruction methods. Current research has made great achievements in the laboratory; however, dealing with rapidly changing light sources, illumination, and imaging parameters in an open environment presents significant challenges for spectral reconstruction. This is because a spectral reconstruction model established under one set of imaging conditions is not suitable for use under different imaging conditions. To deal with the challenges, we explored the feasibility of spectral reconstruction based on camera raw response prediction in this study. In the proposed method, the camera raw response of the training dataset under specific imaging conditions is first predicted via the camera imaging model, then the spectral reconstruction algorithm is applied to spectrally characterize the digital camera based on the training dataset, at last, the spectral reflectance of the testing target is reconstructed from the captured image under the same imaging condition. The study tested the prediction results of raw response values for ColorChecker SG 140 color cards in a closed-light box environment and an outdoor open environment under five different combinations of exposure time and ISO. Spectral reconstruction tests were also conducted using the ColorChecker SG 140 color card for the ColorChecker 24 color card. In the closed lightbox and under different imaging conditions, the average RMSE(%) for reconstructing the true photographic response values using the predicted response values was 4.02, with an average CIEDE2000 color difference of 5.3. In the outdoor open environment and under different imaging conditions, the average RMSE(%) was 3.2, with an average CIEDE2000 color difference of 4.5. The experimental results show that the proposed method still has good reconstruction accuracy in outdoor environments, providing a feasible reference solution for spectral reconstruction in an open environment. In addition, we find that the proposed method is sensitive to the spectral reconstruction algorithms used, and different algorithms have different performances in spectral and chromaticity aspects.
		2025 Vol. 45 (07): 1809-1819 [Abstract] ( 6 ) PDF (14760 KB) ( 2 )

	1820	Multi-Parameter Prediction of Beef Quality Based on Polarized Hyperspectral Imaging
		SONG Ya-fang¹, BU Xiang-tao¹, LI Na², LI Ya-hong^1*, LI De-yang², ZHANG Yun-cui¹, ZHAO Yu³
		DOI: 10.3964/j.issn.1000-0593(2025)07-1820-07
		Because the traditional method of beef quality detection has the disadvantage of destroying samples, it is impossible to carry out multiple tests on the same sample, and the operation is complicated and the test results are lagging, which is difficult to meet the needs of modern food safety detection for non-destructive and rapid response. To solve the problems of complex operation, great destructiveness, and lag of traditional detection methods, polarization hyperspectral imaging detection technology has been widely used in beef quality detection with its advantages of high precision, fast response, and non-destructiveness. This technology can capture the polarization and spectral information of light, not only has the advantages of hyperspectral imaging technology, non-destructive in-depth analysis of the internal quality of beef, at the same time, the technology introduces polarization imaging technology, can inhibit the influence of the surrounding light environment factors on the spectrum of beef, quickly obtain more accurate results, effectively avoiding the shortcomings of traditional methods. This paper used the wavelength range of 900~1 700 nm to compose a near-infrared polarization highlights like technology, combined with high polarization spectral data and a convolutional neural network, to build a more robust quality parameter prediction model for the first time. First, the hyperspectral data of beef samples were collected without polarization and at 0°, 45°, 90°, and 135° polarization angles, respectively, and the regions of interest were extracted. The samples' color parameters (L^, a^, b^) and texture parameters (hardness, adhesiveness, and cohesiveness) were collected. Secondly, the successive projections algorithmwas used to extract the samples' corresponding spectral characteristic wavelengths with unpolarization and different polarization models. Finally, multiple linear regression and convolutional neural network methods were used to construct the multi-parameter prediction model for beef quality. The results show that the prediction accuracy of multiple parameters in polarization mode is better than that without polarization, and the overall prediction effect of the CNN model at 90° polarization Angle is the best. The determination coefficients of parameters L^, a^, b^, hardness, adhesiveness, and cohesiveness were 0.882, 0.905, 0.949, 0.692, 0.671 and 0.911, and the root-mean-square errors of prediction were 0.820, 0.562, 0.461, 3 889.713, 89.746, and 0.027. Compared with unpolarization, the prediction accuracy of the above 6 parameters is at least 13.1% higher, which verifies the feasibility of the convolutional neural network prediction model combined with near-infrared polarization hyperspectral imaging technology in meat nondestructive testing, and provides a new technical idea and method for further meeting the needs of modern food safety testing.
		2025 Vol. 45 (07): 1820-1826 [Abstract] ( 7 ) PDF (8817 KB) ( 7 )

	1827	Intelligent Recognition of Pure Milk Based on Near Infrared Spectroscopy
		HU Shao-wen, HUANG Lang-xin^*, YU Li-hui, WU Zhi-ping, LI Huai-yu, SHI Wei-li, LUO Hong-yu
		DOI: 10.3964/j.issn.1000-0593(2025)07-1827-07
		To prevent the adulteration of pure milk and optimize the detection methods of pure milk, a recognition scheme for pure milk based on near-infrared spectroscopy is proposed in this paper. Firstly, a Fourier near-infrared spectrometer was adopted to obtain the near-infrared spectroscopy signals of different pure milk products from the same brand within the wavelength range of 4 000～10 000 cm^-1. Since the obtained near-infrared spectroscopy signal data is relatively redundant, this paper utilized a principal component analysis algorithm to extract the feature information of the near-infrared spectroscopy signal data within this range to improve the recognition efficiency of pure milk. Four principal components with larger contributions were extracted to obtain the training and testing sample data. Then, the back propagation neural algorithm was employed for preliminary training and testing of the obtained sample data. The test results show that the BP neural network algorithm combined with principal component analysis improves recognition efficiency of pure milk and achieves an accuracy of up to 95%. To further improve the algorithm's accuracy, the particle swarm optimization algorithm was added to the proposed pure milk recognition scheme to optimize the weights and thresholds in the BP neural network. Additionally, a new dynamic decreasing inertia weight factor function was proposed in the particle swarm optimization algorithm for the inertia weight factor. Experiments show that the accuracy of the proposed intelligent recognition scheme for pure milk can be increased to 100%. Therefore, the intelligent recognition scheme for pure milk based on near-infrared spectroscopy can accurately and effectively identify the types of pure milk.
		2025 Vol. 45 (07): 1827-1833 [Abstract] ( 9 ) PDF (6647 KB) ( 3 )

	1834	Research on Emission Spectrum Diagnosis of Laser-Produced Tin Plasma Extreme Ultraviolet Source
		HU Zhen-lin^{1, 2}, WANG Tian-ze^{1, 2}, HE Liang^{1, 2}, LIN Nan^{1, 2*}, LENG Yu-xin^{1, 2}, CHEN Wei-biao³
		DOI: 10.3964/j.issn.1000-0593(2025)07-1834-08
		Due to their characteristics of a small luminous volume, high energy conversion efficiency, high stability, and coherence, laser-produced plasma extreme ultraviolet (LPP-EUV) light sources are widely used in the fields of advanced semiconductor manufacturing and inspection, material surface analysis, and EUV metrology. In this work, the emission spectrum diagnosis of a one μm solid-state laser Sn plasma EUV light source was carried out. First, the 13.5 nm in-band radiation energy, the emission spectrum of the 7~24 nm EUV band, and the 350~750 nm visible light (VIS) band of 1 μm laser-excited solid Sn target plasma in vacuum under different laser peak power densities were measured. The energy conversion efficiency (CE) and spectral purity (SP) were calculated, and the influence of laser peak power density on EUV and VIS spectrum, CE, and SP of Sn plasma was analyzed. Within the parameter range of this experiment, CE increases rapidly at first and then decreases slowly with the increase in laser peak power density, reaching a maximum value of 2.47% at a laser peak power density of 5.2×10¹¹ W·cm^-2. SP increases with the increase of laser peak power density and reaches a maximum value of 7.52% at 1.5×10¹² W·cm^-2. Then, based on the time-resolved VIS spectrum of Sn plasma, the electron temperature (T_e) and electron density (n_e) from 60 to 160 ns after plasma initiation were calculated using the Saha-Boltzmann plot and Stark broadening method, and the temporal evolution of T_e and n_e of Sn plasma in vacuum was studied. The influence of T_e and n_e on EUV band radiation and 13.5 nm in-band radiation was further analyzed. The results show that an increase in laser peak power density leads to an increase in plasma T_e and n_e, and the changes in T_e and n_e affect the distribution of ions with different charge states, causing a change in the EUV radiation spectral distribution. Within the parameter range of this experiment, CE initially increases and then decreases with the increase of T_e and n_e, whereas SP continues to increase. A Te value that is too low will prevent the UTA peak of Sn plasma from reaching 13.5 nm, and a T_e value that is too high will cause more driving laser energy to be converted into EUV radiation below 13.5 nm, resulting in CE not reaching its optimal value. The above research results provide a research foundation and technical support for the engineering development of solid-state laser-driven LPP-EUV light source, as well as the independent development of EUV lithography, EUV metrology, and inspection in China.
		2025 Vol. 45 (07): 1834-1841 [Abstract] ( 8 ) PDF (7334 KB) ( 6 )

	1842	Hyperspectral Identification and Classification of Different Cultivation Methods of A. mongholicus
		WU Qiang¹, LU Ling², FENG Xiao-juan², WANG Meng², WANG Yong-long², HOU Ding-yi³, FAN Bo-bo^2, LIU Jie^2
		DOI: 10.3964/j.issn.1000-0593(2025)07-1842-06
		Huangqi is an important medicinal herb, specifically the dried root of either Astragalus membranaceus (Fisch.) Bge. var. Mongholicus (Bge.) Hsiao (A. mongholicus) or Astragalus membranaceus (Fisch.) Bge. Generally, simulated wild cultivation of A. mongholicus tends to result in higher active compound content compared to horizontal cultivation. However, these differences are challenging to distinguish by visual inspection alone. Traditional methods like High-Performance Liquid Chromatography (HPLC) are accurate for measuring these compounds but are costly and time-consuming. This study aims to develop a rapid, cost-effective, and accurate method to differentiate between simulated wild and horizontally cultivated A. mongholicus. Using a spectroradiometer, we measured the active compound content in ground root samples using HPLC and obtained hyperspectral reflectance data within the 350～2 500 nm wavelength range (SVC-HR1024). The study focused on the spectral characteristics in the visible (VIS, 350～700 nm), near-infrared (NIR, 700～1 100 nm), and shortwave infrared (SWIR, 1 100～2 500 nm) regions. Four machine learning models—Random Forest (RF), K-Nearest Neighbors (KNN), Naive Bayes (NB), and Support Vector Machine (SVM)—were employed for classification. An importance analysis of SHAP features was conducted on the best-performing RF model. The findings reveal that: (1) Simulated wild cultivation had significantly higher active compound content in A. mongholicus than horizontal cultivation; (2) Distinct spectral differences exist between simulated wild and horizontally cultivated A. mongholicus in the NIR and SWIR regions, indicating the impact of the diverse simulated wild environment on pigment synthesis, tissue structure, and chemical composition; (3) The RF model achieved the best performance with an accuracy, precision, F1 score, Kappa, and MCC coefficients of 97.14%, 97.42%, 0.971 3, 0.942 9, and 0.945 6, respectively; (4) SHAP analysis identified key wavelengths associated with moisture, protein, and cellulose content. This study demonstrates the effectiveness of hyperspectral reflectance in distinguishing between simulated wild and horizontally cultivated A. mongholicus samples, providing a novel, non-destructive, and rapid detection method for the quality control and identification of medicinal herbs. This approach has the potential to play a significant role in the quality assessment and market regulation of medicinal herbs.
		2025 Vol. 45 (07): 1842-1847 [Abstract] ( 8 ) PDF (2255 KB) ( 6 )

	1848	A Mechanistic Analysis of Terahertz Absorption Peak Formation in Benzoic Acid and Sorbic Acid Mixtures
		LI Wen-wen, YAN Fang^*, LIU Yang-shuo
		DOI: 10.3964/j.issn.1000-0593(2025)07-1848-09
		This study focuses on benzoic acid and sorbic acid, using a terahertz time-domain spectroscopy (Terahertz Time-Domain Spectroscopy, THz-TDS) system to measure the experimental spectra of their pure substances and mixed pellets in the 0.5~2.2 THz range.After preprocessing the spectral data, the two substances' terahertz (Terahertz, THz) absorption spectra were obtained. The qualitative identification of the above two substances can be accurately achieved based on the peak positions of the characteristic absorption peaks in the absorption spectra. In this paper, the dimer cluster configurations of the benzoic acid and sorbic acid mixture were constructed using the Genmer component. Structural optimizations, frequency calculations, and screening were conducted for their single molecules, unit cells, and mixture clusters using density functional theory (DFT). The potential energy distribution was used to analyze the vibration modes of the characteristic absorption peaks. Combining the simulated spectra with two interaction visualization methods—Independent Gradient Model based on Hirshfeld atom partitioning (IGMH) and the Interaction Region Indicator (IRI)—the vibration modes of abnormal absorption peaks in the mixed pellet were identified. The molecular and intermolecular interaction features were also visualized. The study found that the special absorption peaks in the simulated spectra of the mixture cluster configuration were caused by hydrogen bond vibrations between molecules. This further illustrates that the characteristic absorption peaks of benzoic acid, sorbic acid, and their mixtures in the terahertz frequency range mainly originate from collective vibration modes induced by intermolecular hydrogen bonds and intramolecular chemical bonds. By constructing and screening cluster configurations of mixed systems, this work deeply investigates the causes of characteristic absorption peaks in the terahertz absorption spectra of mixtures. It highlights the impact of intramolecular and intermolecular interactions and hydrogen bond effects on absorption peaks. This study introduces a novel method for characterizing crystal structures, significantly enhancing the accuracy and reliability of spectral interpretation for mixed systems. It also lays a foundation for subsequent quantitative analyses based on spectral data.
		2025 Vol. 45 (07): 1848-1856 [Abstract] ( 5 ) PDF (13940 KB) ( 4 )

	1857	Rapid Detection Method of Bacteria Viability in Water Based on Multi-Wavelength Transmittance Spectra
		HU Yu-xia^{1, 4, 5}, WU Wei-sen¹, ZHANG Rui-xiang^{1, 4}, XUE Fu-rong^{1, 4}, HUANG Shu-long^{1, 4}, SUN Long^{1, 4, 5}, LI Wei-hua³, GAN Ting-ting², ZHAO Nan-jing^2*
		DOI: 10.3964/j.issn.1000-0593(2025)07-1857-09
		Studying the microbial activity of bacteria in water and rapidly assessing the growth status and reproductive capacity of bacteria in water environments are of great significance for the scientific management of water resources and the guarantee of public health security. Although traditional bacterial activity detection techniques, such as the colony counting method, staining observation method and molecular biology methods, are effective, they have cumbersome operation procedures, consume a lot of time and workforce, and are difficult to meet the needs of real-time online detection. In response to the drawbacks of traditional bacterial activity detection techniques, a novel rapid detection method for bacterial activity based on multi-wavelength transmittance spectra is proposed. Taking the common Escherichia coli (E. coli) in water as the research object, the transmittance spectra of E. coli within the wavelength range of 190～800 nm were measured using an ultraviolet-visible spectrophotometer. The characteristics and regularities of the transmittance spectra of E. coli under different proportions of viable bacteria were thoroughly studied. A relationship model between the ratio of double-wavelength optical density and bacterial activity was constructed through the systematic analysis of the correlation and sensitivity of the transmittance spectra of E. coli with changes in activity at different wavelengths. Based on this model, the bacterial activity was calculated, and the accuracy and stability of the calculation results of bacterial activity under different double-wavelength ratio methods were compared and analyzed. The results show that: (1) Due to the differences in the contents of biomolecules in bacteria with different activities, within the wavelength range of 230～300 nm, the bacterial solutions containing higher proportions of viable bacteria have higher optical densities. (2) By analyzing the sensitivity and correlation of the transmittance spectra of E. coli in the wavelength range of 230～300 nm with the activity changes, it was found that the correlation coefficient ranges between the activity of E. coli and the optical densities is 0.959 2～0.993 3, which preliminarily determines the optimal wavelength band for quantitative determination of bacterial activity. (3) Selecting 230 nm as the measurement wavelength and 670 nm as the reference wavelength, a fitting curve of the ratio of optical densities at the two wavelengths to bacterial activity was constructed. The correlation coefficient reached 0.946 2, and the detection range of the viable bacteria proportion was 0%～100%. (4) This fitting curve determined the activities of three E. coli bacterial solutions with different activities. Compared with the plate colony counting method, the maximum relative error was 3.70%, the average relative error was 1.43%, and the accuracy was optimal. This method only requires the optical densities at 230 and 670 nm, and the detection time is within 1 second, which can complete the rapid and accurate detection of bacterial activity. This research achievement provides a new technical idea for the rapid detection and early warning of bacterial activity in water and has potential application value.
		2025 Vol. 45 (07): 1857-1865 [Abstract] ( 10 ) PDF (3415 KB) ( 3 )

	1866	Research on Variety Classification of Starch Based on Terahertz Time-Domain Spectroscopy
		WEI Tao^{1, 2, 3, 4}, WANG Heng^{1, 2, 4}, GE Hong-yi^{1, 2, 4}, JIANG Yu-ying^{1, 2, 5*}, ZHANG Yuan^{1, 2, 4, 5}, WEN Xi-xi^{1, 2, 4}, GUO Chun-yan^{1, 2, 4}
		DOI: 10.3964/j.issn.1000-0593(2025)07-1866-08
		As a major stored carbohydrate, starch is a major source of energy in the human diet and provides more than 50% of the energy needs of the human body. Meanwhile, the starch and its deep-processing industry are fundamental to the national economy and people's livelihoods. However, due to the diversity of starch types and their high similarity in appearance, it is relatively challenging to distinguish amongthem directly. Some illegal merchants often package lower-priced starches as higher-priced starches to increase profits. Consequently, the classification of starch types has significant practical relevance for food processing and industrial production in China. Terahertz (THz) technology, as an effective non-destructive, non-contact, and label-free optical approach, does not produce harmful ionizing radiation during interactions with materials,and can obtain optical parameters such as the absorption coefficient of samples simultaneously. It has a high signal-to-noise ratio and detection sensitivity, and many scholars have applied it to the quality detection of agricultural products. Five of the most common starch samples were selected from cereal starch and rhizome starch to achieve rapid and non-destructive identification of starch. The spectral information was obtained using Terahertz time-domain spectroscopy (THz-TDS) technology, and the absorption coefficient of different starch varieties in the range of 0.2～1.2 THz was calculated based on the experimental data. Subsequently, the original spectra were processed using three preprocessing methods: Savitzky-Golay (S-G) smoothing, multiplicative scatter correction (MSC), and standard normal variate (SNV). Principal component analysis (PCA) was employed to extract feature data based on a cumulative contribution rate exceeding 95%, resulting in the selection of the first three principal components. A multi-classification model was established using the support vector machine (SVM) method. Three types of kernels (linear, polynomial, and radial basis functions) were selected to identify different varieties of starch. The results showed that the PCA-SVM-polynomial combined with SG smoothing achieved the best modeling performance for starch variety classification, with an average accuracy of 0.941 9 on the test set, a Kappa of 0.933, and an F1 score of 0.941 7. Furthermore, this method was compared with logistic regression (LR), decision tree (DT), and random forest (RF). The research results indicated that PCA-SVM was superior to other methods,proving the feasibility of THz technology for starch variety identification and demonstrating important practical application value for the modernization of the food processing industry and the development of starch-based products.
		2025 Vol. 45 (07): 1866-1873 [Abstract] ( 6 ) PDF (6196 KB) ( 5 )

	1874	Research on Defect Detection of GFRP Composites Based on Terahertz Imaging Technology
		ZHANG Yuan^{1, 2, 3, 4}, ZHOU Wen-hui^{1, 2, 3}, GE Hong-yi^{1, 2, 3*}, JIANG Yu-ying^{1, 2, 4}, GUO Chun-yan^{1, 2, 3}, WANG Heng^{1, 2, 3}, WEN Xi-xi^{1, 2, 3}, WANG Yu-xin³
		DOI: 10.3964/j.issn.1000-0593(2025)07-1874-08
		Glass Fiber Reinforced Polymer (GFRP) composites, renowned for their lightweight, impact-resistant, and high-strength properties, have extensive applications in aerospace, automotive manufacturing, and architectural structures. However, the manufacturing process of these composites is often plagued by defects such as pores and cracks, which can severely compromise the material's mechanical strength, leading to product quality degradation and even structural failure, resulting in substantial economic losses for enterprises. This study employs advanced terahertz imaging technology to address the challenge of inspecting epoxy glass fiber composites with various defects. Initially, based on the propagation principle of terahertz waves in transmission mode, a thickness measurement method utilizing time delay difference was adopted to accurately detect and calculate defects at different depths, successfully controlling the error below 0.1 mm, achieving satisfactory detection results. Subsequently, for the quantitative detection of defects with varying sizes, the study converted the original color images of epoxy glass fiber defects into grayscale images, followed by binarization processing using four threshold segmentation methods. Finally, by region labeling, the pixel count of the defective area was calculated, and the defect size was determined by the ratio of defective pixels to total pixels. The results demonstrated that after selecting an appropriate threshold using the manual threshold segmentation method, the root mean square error between the detected area and the actual area could reach 1.368, indicating a close approximation between the detected and actual areas. This experiment confirms that the combination of terahertz imaging technology and image processing methods can quantify the location and size of defects, providing a significant reference for advancing defect detection technology in composite materials. The findings offer new methods and tools for defect detection and quality supervision of other composite materials, holding substantial reference value and enlightening significance, and contributing to enhancing composite product quality. The application of terahertz imaging technology in this study improves the accuracy and reliability of GFRP defect detection and provides a more effective quality supervision approach for the composite material industry. These efforts introduce new ideas and development directions for the future of composite material manufacturing and inspection, promising to drive scientific progress and technological innovation in the field, and exerting a positive impact on industry development.
		2025 Vol. 45 (07): 1874-1881 [Abstract] ( 10 ) PDF (10417 KB) ( 6 )

	1882	An Improved WGAN-GP Generative Adversarial Model in View of NIR Spectral 1^st Derivative Constraint
		LI Zhen-yu¹, ZHAO Peng^{1, 2*}
		DOI: 10.3964/j.issn.1000-0593(2025)07-1882-06
		The Generative Adversarial Network (GAN) has recently become a hot branch in deep neural networks. The mainstream GAN model consists of many improved versions used in image processing and computer vision. These GAN versions are rarely used in spectral analysis. In spectral analysis, they are mainly used to generate synthetic spectral curves so as to extend the classifier's training set for its augmentation and improve its classification performance. Because of the trend of 1D near infrared (NIR) spectral curve, which is an important classification feature and can be quantitatively denoted by a curve derivation, we improve the current one-class Wasserstein GAN with Gradient Penalty (WGAN-GP) model by imposing a spectral 1^st derivative constraint. Specifically, the original NIR spectral vector is connected with the corresponding spectral derivative vector in the revised model L loss function. The concatenated vector is used for model training and spectral curve production. Finally, only the first half is retained in the artificially produced spectral vector to generate the synthetic spectral curves. In our NIR classification experiments of wood species and apple classes, the classification accuracy in some classifiers such as Support Vector Machine (SVM), 1D-Convolutional Neural Network (1D-CNN) and LeNet-5 neural network is increased to some extents after the training set augmentation by use of our improved WGAN-GP model compared with that by use of original WGAN-GP model. Moreover, the NIR spectral curve quality produced by our improved WGAN-GP model has increased greatly, which is indicated by some evaluation measures such as Inception Score, which is computed by use of 1D-CNN instead of the original 2D Inception Net-V3 network, the correlation coefficient between original and synthetic spectral vectors, and these two vectors' difference L₁ and L₂ norms,compared with that by use of the original WGAN-GP model.
		2025 Vol. 45 (07): 1882-1887 [Abstract] ( 7 ) PDF (1617 KB) ( 4 )

	1888	Research on Particle Distribution of DBD Characteristics Based on UV Imaging Technology
		JIANG Song¹, LIU Tong¹, WANG Yong-gang¹, SUN Jiu-ai², WU Zhong-hang², QU Qian^3*
		DOI: 10.3964/j.issn.1000-0593(2025)07-1888-06
		Dielectric barrier discharge (DBD), a method for generating low-temperature plasma, has been widely utilized across various fields. Due to variations in factors such as electrode structure and excitation source, the low-temperature plasma produced by discharge exhibits diverse characteristics, including the type, intensity, and distribution of active particles. Among these, the concentration and distribution of active particles play a crucial role in applying low-temperature plasma. This study aims to investigate the spatial distribution of active particles in the plasma region generated by the discharge. It proposes a diagnostic method for assessing the spatial distribution of active particle concentration, based on ultraviolet spectral imaging technology. Utilizing ultraviolet spectral imaging, a corresponding bandpass filter is employed to convert the collected ultraviolet (UV) spectral image into turbo mapping. The image grayscale is then extracted to analyze the spatial distribution of characteristic particle concentration. This study uses N₂ (337.1 nm) active particles, generated by needle-plate DBD discharge driven by pulse power, as an example to analyze their spatial distribution characteristics under varying parameters. The results indicate that, under pulse voltage drive, N₂ (337.1 nm) excited state particles are primarily distributed along the axis of the discharge channel formed by the needle tip structure and begin to diffuse near the dielectric. The deposition electric field, generated by the charge on the surface of the dielectric plate, intensifies the ionization and collision of nearby particles, resulting in deposition on the surface of the dielectric plate as a large area of high concentration. Along the centerline of the discharge channel, the intensity is highest at the needle tip and gradually decreases as the distance increases. Particle concentration increases on the surface of the dielectric. When the voltage reaches a sufficient level, a second intensity peak appears at the head of the channel, primarily due to the high density of high-energy electrons. In the discharge channel, the particle concentration is predominantly centered in the channel's center and rapidly decays towards both sides, forming a clear boundary. As the voltage increases, the boundary becomes more pronounced. Finally, the Bland-Altman plot method was employed to verify that the method proposed in this study is highly consistent with the particle intensity change trend and amplitude reflected by the emission spectrum, validating the accuracy of the ultraviolet spectral imaging detection method, which can aid in subsequent particle regulation and enhance application efficiency.
		2025 Vol. 45 (07): 1888-1893 [Abstract] ( 8 ) PDF (15688 KB) ( 4 )

	1894	Study on the Factors Affecting the Signal-to-Noise Ratio of Two-Photon Optical Frequency Standard Spectral Lines Based on Rubidium Atom
		ZHANG Jiong-yang¹, ZHAI Hao^{1, 2}, XIAO Yu-hua^1, WANG Ji^1*, DAI Hu¹, CHEN Jiang¹
		DOI: 10.3964/j.issn.1000-0593(2025)07-1894-06
		The optical frequency standard based on two-photon transition is expected to become a miniaturized optical atomic clock available in the future. Acquiring spectral lines with a high signal-to-noise ratio is an important prerequisite for achieving high-performance optical frequency standards. The experimental setup of the two-photon transition optical frequency standard was completed. The factors affecting the signal-to-noise ratio of the optical frequency standard spectral line were experimentally analyzed from the aspects of laser intensity, atomic density, and the gain of the photomultiplier tube, and the signal-to-noise ratio of the optical frequency standard transition spectral line under different parameter values was obtained. Experimental results show that the spectral line's signal-to-noise ratio increases linearly with the increase of the laser intensity in the range of 15 700 mW·mm^-2.When the atomicitydensity reaches 1.5×10¹³ cm^-3 and the PMT gain is 1.2×10⁵, the signal-to-noise ratio of the spectral line reaches saturation. The signal-to-noise ratio of the spectral line of the two-photon transition optical frequency standard obtained by this setup can reach up to 2600. Considering that the linewidth of the two-photon transition spectral line is 1MHz, the short-term frequency stability of the optical frequency standard is expected to reach 3×10^-13τ^-1/2. By studying the factors affecting the signal-to-noise ratio of two-photon optical frequency standard spectral lines based on Rubidium atoms, the atomic transition spectral line with a high signal-to-noise ratio is obtained, which is of great significance for the development of high-performance and integrated two-photon optical frequency standards.
		2025 Vol. 45 (07): 1894-1899 [Abstract] ( 6 ) PDF (3427 KB) ( 4 )

	1900	Rapid Measurement of Vitamin E in Vitamin E/Ultra-High Molecular Weight Polyethylene Powders by Infrared Spectroscopy
		HU Zhi-jie¹, LU Man-li^2*, TIAN Ji-li¹, ZHANG Wen-li², WANG Mou-hua²
		DOI: 10.3964/j.issn.1000-0593(2025)07-1900-06
		Vitamin E doped modified ultra-high molecular weight polyethylene (VE/UHMWPE) has important applications in the medical field. Artificial joints formed by compression molding of VE/UHMWPE powder have been clinically used, and the antioxidant properties of VE have extended the implantation time of this type of artificial joint in vivo. Therefore, studying the amount of VE added in VE/UHMWPE powder has crucial guiding significance for protecting artificial joints. This paper investigated a method for rapid quantitative analysis of VE in VE/UHMWPE powders by Fourier transform infrared spectroscopy (FTIR).Five VE/UHMWPE powders with different VE concentrations were made into discs of different thicknesses using the hot press method, and infrared testing was performed in transmission mode. All spectral data of the samples were analyzed and studied by OMNIC. The characteristic peaks of VE located at 1 210 and 1 260 cm^-1 were used as target peaks, and the absorption peaks at 2 020 and 1 360 cm^-1 in UHMWPE were used as reference peaks, calculating I_{1 210/2 020}, I_{1 260/2 020}, I_{1 210/1 360} and I_{1 260/1 360}, respectively, and statistically analyzed. The precision and repeatability of results were determined using relative standard deviation (RSD). The results indicated that: sample preparation with 40~80 mg of powder gave the best results, with RSD values <5% in the repeatability test; the I_{1 210/1 360} index was used as the evaluation parameter for quantitative analysis with the best accuracy and RSD value <3%. The method provides a reference for the application of rapid nondestructive testing of the actual VE content in VE/UHMWPE powders.
		2025 Vol. 45 (07): 1900-1905 [Abstract] ( 7 ) PDF (5677 KB) ( 4 )

	1906	Study on Modeling the Effect of Three-Dimensional Fluorescence Spectrum of Predicting the Content of Peanut Oil Adulterated Soybean Oil
		WEI Quan-zeng¹, LIU Xue-ying¹, WANG Zhi-jie¹, DING Fang²
		DOI: 10.3964/j.issn.1000-0593(2025)07-1906-10
		To determine the content of adulterated peanut oil in soybean oil, the three-dimensional fluorescence spectrum data of soybean oil and peanut oil counterfeit were collected. Rayleigh scattering and Raman scattering were removed using the triangular internal interpolation method. Then the fluorescence spectra were processed using Savitzky-Golay. The Alternating trilinear decomposition (ATLD) and Parallel factor (PARAFAC) algorithms were used to predict peanut oil content. Meanwhile, after scattering and smoothing the three-dimensional fluorescence data of the different contents of counterfeit peanut oil. The emission spectrum corresponding to each excitation wavelength is decomposed by wavelet packet decomposition (WPD), and the wave packet coefficient of the lowest frequency band is used as the characteristic amount of fluorescence emission spectrum data. All the emission wavelengths were reconstructed according to the sequence number of excitation wavelengths, and the data were reconstructed into a first-order fluorescence spectrum data vector. Partial least squares (PLS) and artificial neural network (ANN) data models were constructed to predict the content of peanut oil in counterfeit products. The results indicated the regression coefficients R² of PARAFAC, ATLD, WPD-PLS, and WPD-ANN were 0.898, 0.941, 0.961, and 0.981, respectively. Mean absolute deviation (MAD), mean squared error (MSE), and root mean squared error (RMSE) of the training set, verification set, test set, and all data of the WPD-ANN algorithm model were all small. The peanut oil content in counterfeit products was predicted using the WPD-ANN model. The percentage of samples with prediction deviation within ±5% was 82.5%. The peanut oil content prediction results by WPD-ANN, WPD-PLS, ATLD, and PARAFAC were compared and analyzed. The mean and median deviations of WPD-ANN and WPD-PLS models are near 0%, while the mean and median deviations of ATLD and PARAFAC models are far from 0%. Compared with the PARAFAC model, the ATLD model has faster convergence and smaller deviation. ATLD and PARAFAC models may be affected by nonlinear factors, and their prediction effect was inferior to that of WPD-ANN and WPD-PLS, while ANN and PLS were based on first-order data regression modeling after WPD and data reconstruction. ANN was a nonlinear model. Therefore, the WPD-ANN model has stronger prediction ability and smaller deviation for peanut oil content in counterfeit peanut oil. The WPD-ANN model was the best algorithm among the four algorithms for predicting peanut oil content in counterfeit peanut oil. This provides a research basis for quantitative analysis of adulterated edible oil.
		2025 Vol. 45 (07): 1906-1915 [Abstract] ( 9 ) PDF (13303 KB) ( 3 )

	1916	Rapid Detection for Xylose Content Using Near-Infrared Spectroscopy Technology
		LAN Xi-hua, WANG Zhi-guo^*, LUAN Xiao-li, LIU Fei
		DOI: 10.3964/j.issn.1000-0593(2025)07-1916-08
		Xylose, as a functional oligosaccharide, possesses health benefits such as antioxidant properties and promoting intestinal health, and is widely used in food, medicine, and biofuels. There is still a lack of effective rapid detection methods for xylose content. An online detection method based on near-infrared spectroscopy technology is proposed to address the issue of content detection during xylose production. Firstly, sample solutions are collected and scanned using a near-infrared spectrometer to obtain raw spectra. The raw spectra are then preprocessed using first derivative and smoothing filter methods to remove noise and baseline drift effects. Subsequently, the random frog algorithm is employed for feature selection of spectral variables, and the prediction relative analysis error is used to search for the optimal number of features. The results show that the model's predictive performance is optimal when the number of features is between 20 and 30. Considering other indicators, the number of features is selected as 25, determining the wavelength characteristics representing xylose content. Due to the random subset selection and random forest regression characteristics of the random frog algorithm, this algorithm has obvious advantages in performing the task of feature wavelength screening for high-dimensional xylose data, but also has the defect of low result reproducibility. After obtaining the wavelength features, the results are weighted and accumulated to weaken the impact of the algorithm's uncertainty on the final model. Then, a predictive model for xylose content is established using data measured by a liquid chromatograph as labels. Finally, the method is used to rapidly determine the xylose content of samples collected from the process site, and the prediction effects are compared with those of the PLS and Lasso models. The results indicate that the training set determination coefficient R²=0.937 7, and the test set determination coefficient R²_p=0.933 5, with R² and R²_p close to 1, indicating that the model can explain the training set data well and has good generalization performance. The prediction root mean square error RMSEP=5.844 6, and the prediction relative analysis error RPD=3.879 2>2.5, indicating that the model can predict the xylose content of samples relatively accurately. Through comparison, it is found that the RJFA-PLS model's evaluation indicators are superior to those of the PLS model, with RMSEP reduced by 112.7%, and R², RPD, and R²_p increased by 21.8%, 52.5%, and 24.6%, respectively. However, the Lasso algorithm performs poorly predicting xylose content based on this dataset. Under the experimental conditions of this study, the model established using the above method is more suitable for predicting xylose content than the PLS and Lasso models. The proposal of this method solves the problem of lag in xylose content detection results and also provides a prerequisite for the research of online detection technology for xylose.
		2025 Vol. 45 (07): 1916-1923 [Abstract] ( 6 ) PDF (2562 KB) ( 3 )

	1924	Rapid Determination of Moisture Content of Freeze-Dried Carrots by Terahertz Spectroscopy Combined With Machine Learning Algorithms
		SUN Meng¹, CHENG Jun¹, DIAO Shu¹, HAN Tian-yu², YU Zhi-long¹, LI Jing-wen², XIE Yun-fei^1*
		DOI: 10.3964/j.issn.1000-0593(2025)07-1924-08
		Moisture content (MC) is vital to freeze-dried carrots' quality and shelf life. However, traditional moisture measurement methods are time-consuming and inefficient. Therefore, this study aimed to develop a rapid, nondestructive detection method utilizing terahertz time-domain spectroscopy (THz-TDS) and machine learning (ML) technology to determine the moisture content of freeze-dried carrots. The time-domain spectral data for 140 samples with varying moisture content were collected. Based on the optical parameter extraction model, the' absorption coefficient spectrum and refractive index spectrum of these samples within the terahertz frequency band were obtained. To enhance the quality of the spectral data, the acquired spectra underwent preprocessing through moving average (MA) smoothing and Savitzky-Golay (SG) smoothing. Subsequently, three feature extraction algorithms: competitive adaptive reweighting sampling (CARS), successive projection algorithm (SPA), and uninformative variable elimination (UVE), were employed to filter out the spectral variables most closely related to water content from the original spectral data. Finally, three machine learning algorithms: partial least squares regression (PLSR), back propagation artificial neural networks (BPANN), and extreme gradient boosting (XGBoost) were utilized to construct quantitative prediction models. These models were then comprehensively evaluated using model evaluation indices to determine the optimal optical parameters and the most effective algorithm combination for detecting the moisture content of freeze-dried carrots. The results indicated that the absorption coefficient spectrum accurately and effectively captured the moisture information. Pretreatment effectively reduced spectral noise, and feature extraction identified the key variables related to moisture. BPANN exhibited the best quantitative prediction performance among the machine learning algorithms tested. Specifically, the SG-CARS-BPANN model, which was based on the absorption coefficient spectrum, demonstrated the strongest predictive capability (R²_C=0.971 2，RMSEC=0.007 3，R²_P=0.936 6，RMSEP=0.010 7). These findings demonstrated that the combination of THz-TDS and machine learning algorithms can realize rapid and nondestructive moisture detection in freeze-dried carrots, and the established method has the potential to monitor moisture content in freeze-dried fruits and vegetables in real time during drying and storage.
		2025 Vol. 45 (07): 1924-1931 [Abstract] ( 5 ) PDF (10761 KB) ( 3 )

	1932	Determination of Soluble Solids Content in Fresh Corn by Near Infrared Spectroscopy
		YANG Guang-hui^{1, 3}, ZHANG Yong-li^{1, 2, 3}, WANG Mei-pan^{1, 3}, LIU Yan-de⁴, JIANG Xiao-gang⁴, SUN Jing^{2, 3}, ZHOU Xin-qun^{2, 3}, HAN Tai-lin^1
		DOI: 10.3964/j.issn.1000-0593(2025)07-1932-08
		China is the world's largest producer and consumer of fresh corn. Soluble solids content (SSC) is a key indicator of the quality of fresh corn, and there is an urgent need for effective and rapid non-destructive testing methods to respond to the market demand for the test. In order to realize the rapid and nondestructive detection of fresh corn SSC, a prediction model of fresh corn SSC based on near-infrared spectral features combined with a chemometrics method is proposed. Taking sweet corn as the research object, using the near-infrared (NIR) detection device built independently by the laboratory, we explored the multi-point acquisition method based on fresh corn materials to obtain the NIR diffuse reflectance spectra in the middle of the cob, and after the anomalous spectra were excluded by the Mahalanobis distance method, 103 samples were selected for modeling. The dataset is divided into training and test sets according to the ratio of 4∶1, and five algorithms, including Savitzky-Golay smoothing (SGS), Standard Normal Transform (SNV), Multivariate Scattering Correction (MSC), First-Order Derivative (FD), and De-Trending (DT), are applied to preprocess the spectral data and build the SSC full-band prediction model. The competitive adaptive reweighting algorithm (CARS), successive projection algorithm (SPA), and random frog hopping algorithm (RF) are used for feature band selection, and the SSC feature band model based on the partial least squares regression algorithm (PLSR) and support vector machine regression algorithm (SVR) is established. The results show that: SNV, MSC, and FD achieved better preprocessing results, and the prediction accuracy of the eigen-band model was significantly improved compared with the full-band modeling. The “SNV-CARS-PLSR” model built by SNV preprocessing combined with CARS feature extraction performs optimally, The training set coefficient of determination (R²_C), training set root mean square error (RMSE_C), test set coefficient of determination (R²_P), test set root mean square error (RMSE_P), and residual prediction deviation (RPD) were 0.869, 0.219, 0.858, 0.191, and 2.715, respectively. Compared to the SNV-preprocessed full-band model, the “SNV-CARS-PLSR” model improves the R²_P of the test set by 12.3%. Comparing the different feature band modeling methods, the “SNV-CARS-SVR” model based on SVR is slightly better than the “SNV-CARS-PLSR” model based on PLSR. The “SNV-CARS-SVR” model has an R²_C of 0.881, an RMSE_C of 0.207, an R²_P of 0.869, an RMSE_P of 0.185 and an RPD of 2.843. This study can provide technical support for rapidly detecting SSC in fresh corn cobs based on near-infrared spectroscopy.
		2025 Vol. 45 (07): 1932-1939 [Abstract] ( 8 ) PDF (3897 KB) ( 4 )

	1940	Pathogenic Bacteria Raman Spectrum Classification Method Based on Diffusion Kernel Attention
		WU Shu-lei^{1, 2}, ZHANG Jia-tian^{1, 2}, WANG Jia-jun², DANG Shi-jie², ZHAO Ling-xiao², CHEN Yi-bo^2*
		DOI: 10.3964/j.issn.1000-0593(2025)07-1940-06
		Current methods for identifying pathogenic bacteria are time-consuming, leading to delays in optimal treatment and promoting antibiotic resistance. Therefore, developing a rapid, accurate, culture-free technique for this scenario has high clinical value. Raman spectroscopy can serve as a molecular fingerprint for rapid bacterial species identification, and computer-assisted classification is the current research hot spot. However, the classification methods based on machine learning and CNN in related works have poor generalization and insufficient feature mining ability, which leads to low classification accuracy. This study innovatively proposes a deep learning network named Raman Transformer (RaTR). RaTR can improve feature miningcapability and classification accuracy using kernel attention computation based on radial basis kernel function, and its model generalization is enhanced by introducing the diffusion process. Moreover, the discrete wavelet transform is proposed to address the excessive parameters and few-shot issues. Experimental validation on the Bacteria_ID and ATCC datasets shows that RaTR achieves classification accuracies of 85.83% and 84.73% respectively, demonstrating its accuracy and strong generalization. Visualizing key spectral features further confirms the effectiveness of feature extraction by the model. Finally, visualizing the spectral key features further confirms the effectiveness of RaTR feature extraction.
		2025 Vol. 45 (07): 1940-1945 [Abstract] ( 3 ) PDF (14612 KB) ( 2 )

	1946	K-Means-CNN-Based Classification Study of Mixed Alloy Samples of Complex Grades
		MA Yao-an¹, HUANG Yu-ting¹, ZHANG Jian-hao¹, QU Dong-ming¹, HU Bei-bei², LIU Bi-ye², YANG Guang¹, SUN Hui-hui^1*
		DOI: 10.3964/j.issn.1000-0593(2025)07-1946-07
		Laser-induced breakdown spectroscopy (LIBS) is a highly efficient elemental analysis method with simple sample preparation, non-contact measurement, strong field adaptability, and fast analysis speed by focusing an ultra short pulse of laser light on the surface of the sample to form a plasma, and then analyzing the emission spectrum of the plasma to determine the material composition and content of the sample. Using LIBS technology for elemental analysis, component classification, and identification is the key direction of the research. At present, LIBS technology is mainly used in rock and mineral detection, environmental monitoring, chemical identification, and related fields, while less research is conducted on the classification of mixed alloys with multiple components and complex grades. The commonly used high-performance, accurate classification algorithms usually require high computational resources and are difficult to mount on portable and miniaturized LIBS systems. A mixed sample of various grades of AL, FE, and CU alloys was excited by an MPL-T-1064 laser with a modulated optical path through a front mirror set to collect data. Data were preprocessed using Principal Component Analysis (PCA) and then input into the K-means clustering algorithm (K-means), a Convolutional Neural Network (CNN) model for classification. The K-means is unable to classify complex alloy grades finely, but has an accuracy of 99.97% in the work of large class differentiation.CNN can classify complex alloy grades finely with an accuracy of 99.15%, but it has a relatively high demand on computational resources. Aiming at the above problems, a fusion algorithm is designed to use the K-means algorithm to process the mixed alloy spectral data. It coarsely classifies samples of the same kind but different grades. Then, the data after the first-stage classificationis input into the CNN model to carry out fine classification. The accuracy of classification in the mixed alloy spectra of ten kinds of samples of grades of AL, FE, CU reaches 99.35%, and the accuracy in the 5-fold cross-validation reaches 99.52%, which verifies that the algorithm has better generalization ability while classifying accurately. The classification accuracy of the fusion algorithm is 39.65% higher than that of the K-means algorithm, and the running speed is 21.94% faster than that of the CNN algorithm. It provides an efficient, fast, and accurate method for the classification of mixed alloys with multiple compositions and complex grades. It provides a new idea for developing a more lightweight and portable LIBS system.
		2025 Vol. 45 (07): 1946-1952 [Abstract] ( 5 ) PDF (4744 KB) ( 3 )

	1953	The Scientific Analysis of Materials of Polychrome Paintings From Northeast Chonglou of the Forbidden City
		JIANG Yi¹, PAN Jiao¹, DUAN Hong-ying^2*
		DOI: 10.3964/j.issn.1000-0593(2025)07-1953-08
		The Northeast Chonglou is located in the Outer Court area of the Forbidden City and is a unique building among the three main halls of the Forbidden City. It is an important component of the outer court architectural complex. The Northeast Chonglou was first built during the Ming Dynasty Yongle period, but was later rebuilt and renovated after being affected by the fire. The existing architectural polychrome paintings on the inner eaves belong to the early Qing Dynasty YawumoXuanzi polychrome paintings, which are well preserved and an important sample for studying early Qing Dynasty polychrome paintings. This study sampled 10 architectural polychrome paintings and 1 red oil decoration from Northeast Chonglou. Using various analytical instruments such as a metallographic microscope, a laser Raman spectrometer, a scanning electron microscope-energy spectrometer, an X-ray diffractometer, and a pyrolysis-gas chromatography/mass spectrometry to conduct scientific analysis on the samples. The results indicate that the polychrome painting conforms to the characteristics of the early Qing Dynasty. In combination with the findings from literature review, there has been no repainting or restoration of the polychrome paintings in the later period; The pigments used for polychrome painting are copper trihydroxychlorides, malachite, lead white, chalk, organic pigment indigo, and carbon black; The green pigment is obtained by mixing copper trihydroxychlorides with malachite; The white pigment is a mixture of lead white and chalk. The red pigments are cinnabar and hematite, and hematite plays the role of interlining color under vermilion. The plaster layer uses brick ash as an inorganic filling material, mixed with cooked tung oil to cover the surface of the wood. Then the pigment mixed with cooked tung oil was painted on the plaster layer. The drawing process follows the sequence of Qing Dynasty style polychrome painting from dark to light colors. This study is the first comprehensive analysis of Northeast Chonglou. It provides basic information for its later conservation and restoration,as well as valuable materials for studying the official polychrome painting of the early Qing Dynasty.
		2025 Vol. 45 (07): 1953-1960 [Abstract] ( 6 ) PDF (27933 KB) ( 3 )

	1961	The Heat Treatment Process and Color Mechanism of Green Fluor-Hydroxyapatite
		YUE Su-wei^1, YAN Xiao-xu^2, CHEN Si-min², LUO Jie²
		DOI: 10.3964/j.issn.1000-0593(2025)07-1961-07
		Apatite is widely used as an inorganic non-metallic material, a geological indicator mineral, and a common gemstone variety. The chemical formula for apatite is [Ca₁₀(PO₄)₆(F, Cl, OH)₂], Ca²⁺sites can be isomorphously substituted by divalent transition metal cations (Fe²⁺, Mn²⁺, etc. ), rare earth element ions (La³⁺, Ce³⁺, Gd³⁺, etc. ), or alkali and alkaline earth metal ions (K⁺, Na⁺, Sr²⁺, etc. ), resulting in various colors such as yellow, green, blue, purple, and photochromic effects. Green fluor-hydroxyapatite turns into a bright blue hue after heat treatment at 650～700 ℃, possessing high ornamental and commercial value. The results show that the samplesare mainly composed of Ca²⁺ and P⁵⁺ with a small amount of N⁵⁺, which is mainly substituted [PO₃] in the form of —NH₂, indicating biochemical behavior correlated in formation. UV-Vis results show that the sample's color is mainly caused by the isomorphic substitution of Ce³⁺ and Mn²⁺ in the distorted octahedron of Ca²⁺, forming a blue-green transmission window. The defects in the green apatite sample before heat treatment are mainly ionic centers F^--O^--F^-, combining with O^2--O^--V and OH^--O^--V, caused by the isomorphic substitution of Ce³⁺ at the CaⅠ site. After heat treatment, the recovery of hole centers is mainly manifested as ionic center defects caused by Mn²⁺ substituting the distorted octahedron at CaⅠ and CaⅡ sites. Combining EPR spectra shows that Mn²⁺ can be distributed at both CaⅠ and CaⅡ positions, causing a blue hue in the samples; otherwise, when Ce³⁺ is highly substituted at the CaⅠ position, the samples turn green.
		2025 Vol. 45 (07): 1961-1967 [Abstract] ( 7 ) PDF (7424 KB) ( 10 )

	1968	Study of Near-Infrared Fingerprints of Ganoderma Lucidum in Different Growth Environments
		TAN Fang-ping¹, LU Tong-suo^{1, 2*}
		DOI: 10.3964/j.issn.1000-0593(2025)07-1968-11
		Ganoderma lucidum (G. lucidum), a precious fungus with a long history of medicinal use, faces challenges in authenticity identification and quality evaluation due to its diverse species and varying growth environments. This study integrates near-infrared spectroscopy (NIRS)and spectral preprocessing methods to analyze G. lucidum samples from different growthenvironments in the market. By collecting and analyzing NIRS data of five G. lucidum samples, spectral preprocessing methods—including baseline correction (first-/second-order derivatives, continuous wavelet transform (CWT)) and scattering correction (multiplicative scatter correction (MSC) , standard normal variate transformation (SNV))—were applied to eliminate background noise. This enhanced the accuracy of spectral data in reflecting the intrinsic characteristics of G. lucidum, with a focus on the distribution of characteristic absorption peaks and hydrogen-containing groups in the NIRS region. The results revealed significant differences in NIRS between the stipes and pilei of G. lucidum from distinct growth environments, exhibiting unique characteristic absorption peaks and hydrogen-containing group distributions. These absorption peaks were closely associated with active components in G. lucidum, where environmental factors (e. g., temperature, humidity, light intensity) and cultivation conditions (e. g., substrates such as sawdust, wheat bran, and gypsum) were identified as key determinants of active component synthesis and distribution. Environmental factors influence growth cycles and metabolic activities, while cultivation substrates affect growth rates, mycelial vigor, biomass, and morphological parameters (stipe length, stipe diameter, pileus diameter, pileus thickness). This method provides an effective approach for the authenticity identification of G. lucidum. Its implementation significantly enhances quality control of G. lucidum products, facilitates comprehensive quality evaluation, aids in eliminating inferior products in the market, and ensures consumers select appropriate G. lucidum varieties to meet market demands.
		2025 Vol. 45 (07): 1968-1978 [Abstract] ( 8 ) PDF (27729 KB) ( 3 )

	1979	XPS Depth Profile Analysis of Sodium Molybdate-Based Conversion Film on Bronze Surface
		FAN Chen-xiao, LI Chang-qing, LUO Yu-jia, HE Bei, YANG Ben^, JIN Pu-jun^
		DOI: 10.3964/j.issn.1000-0593(2025)07-1979-07
		Inorganic molybdate corrosion inhibitors have broad application prospects in conserving metal cultural relics due to their environmentally friendly, low-toxicity, and high-efficiency properties. This study used sodium molybdate solutions of varying concentrations to construct corrosion-inhibiting conversion films on bronze surfaces via chemical deposition. These films' composition, structure, and performance were systematically investigated using electrochemical testing and XPS depth profiling. Experimental results showed that after 1 day of immersion, the bronze sample treated with 0.2 mol·L^-1 sodium molybdate solution exhibited a relatively high corrosion inhibition efficiency of approximately 50%. In contrast, samples treated with lower concentrations (0.02 and 0.05 mol·L^-1) demonstrated lower efficiencies. With prolonged immersion, the inhibition efficiency of samples treated with 0.2 and 0.5 mol·L^-1 solutions gradually decreased, while those treated with 0.02 and 0.05 mol·L^-1 solutions initially increased before subsequently decreasing.XPS depth profiling analysis revealed that within the 1.25～5.00 μm depth range, redox reactions occurred, forming metal oxides such as SnO₂, CuO, Cu₂O, and MoO₂. Specifically, Cu exhibited a layered distribution, transitioning from an outer Cu₂O layer to an intermediate Cu₂O+CuO transition layer and finally to an inner CuO layer. Mo existed predominantly as MoO^2-₄ in the outer film and gradually converted into a mixed form of MoO^2-₄ and MoO₂ in the inner layers, with molybdate ions being reduced to MoO₂. Additionally, treated samples exhibited a noticeable color change due to the molybdenum blue phenomenon. After 3 days of immersion, the ΔE^*_Lab color difference of the sample treated with 0.2 mol·L^-1 sodium molybdate solution was approximately 25, whereas that of the sample treated with 0.5 mol·L^-1 solution was about 46.This study elucidates the chemical composition and structural characteristics of molybdate-based corrosion-inhibiting conversion films on bronze surfaces, providing valuable insights into their application in conserving metal cultural relics.
		2025 Vol. 45 (07): 1979-1985 [Abstract] ( 9 ) PDF (10325 KB) ( 4 )

	1986	Design of Transient Temperature Field Detection System for Fire Explosion
		XIAO Ju¹, HAO Zhi-yong¹, ZHOU Man-lan¹, HU Wei-zhao²
		DOI: 10.3964/j.issn.1000-0593(2025)07-1986-06
		The transient distribution of the temperature field is a crucial indicator for assessing the extent of explosion damage. Explosions, detonations, and other phenomena that occur during a fire can pose significant risks to emergency rescue operations. It is necessary to study the changes in the transient temperature field during the explosion process and the extent of damage. It can improve the safety of fire extinguishing. The distribution range and temperature values of the transient temperature field during the explosion process were quantitatively analyzed. A radiation temperature measurement system based on spectral normalization was designed. The existing literature primarily employs radiation thermometry to measure the temperatures of flames from explosions. Most literature uses a single wavelength to calculate the brightness temperature field within the explosion zone. But this method cannot calculate the true temperature value. A structure for image acquisition based on a multi-wavelength combination narrowband filter partition has been designed. A spectral normalization radiometric temperature measurement algorithm was proposed. The system consists of an imaging lens group, a multi-wavelength combination narrowband filter, and a multispectral camera. The imaging lens group is used to collect explosive radiation from the test area and achieve collimation and focusing. A multi-wavelength combination narrowband filter is a combination of four narrowband filters with different characteristic wavelengths. It enables simultaneous acquisition of images on the photosensitive surface of the CCD array. This design sacrifices 1/4 of the spatial resolution in exchange for simultaneously obtaining test images of four characteristic wavelengths. Finally, the multispectral camera simultaneously captures multispectral images under four characteristic wavelength conditions. The processing module is used to complete temperature inversion based on radiance. Finally, obtain the temperature field and transient changes within the explosion area. The experiment utilized the S16 thermocouple sensor to calibrate the instantaneous temperature at the actual location of the explosion area, and the M20 infrared thermal imager's test results were used to calibrate the transient temperature range in the explosion area. The temperature test results of the thermocouple sensor show that the highest temperatures at distances of 1.0 and 5.0 m are 1 625 ℃ and 810 ℃, respectively. The inversion results of this system are 1 602 ℃ and 783 ℃, respectively, with an average relative error of 3.1%. The range test results of the thermal imager show that the maximum range is 6.9 m×6.0 m. This system is 7.2 m×6.3 m, with an average relative error of less than 5%. It verifies the feasibility of using four characteristic wavelength partition images to invert the instantaneous temperature field of the explosion area in this system. The experimental results demonstrate that the transient temperature inversion accuracy of this system is high, enabling it to reconstruct the three-dimensional temperature field. This design can dynamically identify the explosion range. It has greater potential and practical value in fields such as fire and explosion.
		2025 Vol. 45 (07): 1986-1991 [Abstract] ( 7 ) PDF (24181 KB) ( 3 )

	1992	Comparison of Wavelength Screening Methods for Insulator Pollution Hyperspectral Detection
		LI Jia-jia¹, WANG Xiang-feng¹, HUANG Fei-lin¹, LIU Yong¹, YAO Xuan², YANG Hui², CHENG Hong-bo^2*
		DOI: 10.3964/j.issn.1000-0593(2025)07-1992-07
		The surface pollution will affect the insulator's insulation ability and harm the power system. The current detection method requires power outage sampling, which is tedious and time-consuming. Hyperspectral analysis can realize non-contact non-power outage detection and has good potential for application in insulator pollution detection. In order to reduce the data processing amount of insulator surface pollution hyperspectral detection and improve the accuracy of insulator hyperspectral data classification and identification, according to the current manual test sample preparation standards, Three artificial pollution samples of insulators with salt density of 0.22 milligrams per square centimeter, ash density of 0.1 milligrams per square centimeter and salt density of 0.3 milligrams per square centimeter and ash density of 0.1 milligrams per square centimeter were made.Hyperspectral sampling was carried out on insulator samples of different pollution levels. 15 regions were selected on each sample to extract regional spectral data, and a total of 90 groups of spectral data were obtained. 63 training set samples and 27 test set samples were selected. Competitive adaptive reweighted sampling (CARS) Algorithm, Successive Projections Algorithm (SPA and Uninformation Variables Elimination (UVE) were used to screen the characteristic wavelengths of insulator hyperspectral data and build a support vector machine classification model. Multivariate scattering correction (MSC), standard normal variation (SNV), first order derivative, deconvolution, moving average filtering, baseline correction, normalization, and wavelet transform were used to preprocess the spectral data. The classification experiments were carried out using the tested sample data, and the classification effects of different pretreatment methods and feature wavelength screening methods were compared. The experimental results show that the pre-treated data can improve classification recognition accuracy, and the pre-treated data's recognition accuracy can reach a low of 51.85% and a high of 96%, higher than the 40.74% in the untreated condition. Feature screening can reduce the dimensionality of the original spectral data, and SPA is the most efficient of the three methods, with an average screening rate of 3.56%. After screening, the accuracy of classification recognition can be improved, and the accuracy of classification recognition of original data can be increased from 40.74% to 74.07% after CARS screening. Pretreatment of spectral data combined with feature wavelength screening can greatly improve the classification and identification accuracy of dirty insulators. The classification and identification accuracy of MSC-CARS, SNV-CARS, MSC-UVE, and normalized UVE can reach 100% after combined processing. The average classification accuracy of CARS, SPA, and UVE combined with 8 preprocessed test sets was 87.05%, 86.25%, and 83.47%, respectively, indicating that the combination of preprocessing and feature screening plays an important role in improving data quality and model performance. These preprocessing and characteristic wavelength screening methods of spectral data can effectively reduce the original data dimension and simplify the model complexity, which can play an important role in improving the accuracy of insulator pollution classification and recognition.
		2025 Vol. 45 (07): 1992-1998 [Abstract] ( 6 ) PDF (6447 KB) ( 3 )

	1999	The Effect of Laser Incidence Angle on the Signal Intensity and Repeatability of High-Frequency Laser-Induced Breakdown Spectra
		YANG Wen-feng¹, XIE Min-yue¹, QIAN Zi-ran¹, LI Shao-long¹, CAO Yu², LIN De-hui¹, LÜ Shuang-qi¹, YANG Xiao-qiang¹, HUA Hai-jie¹
		DOI: 10.3964/j.issn.1000-0593(2025)07-1999-09
		In-situ, real-time online monitoring technology helps improve the controllability of the aircraft skin removal process. Laser Induced Breakdown Spectroscopy (LIBS) is a highly efficient and fast elemental analysis technique, which realizes intelligent laser-controlled paint removal by quickly analyzing the elemental changes in the paint layer through the plasma spectral information generated during the laser's interaction with the material. LIBS technology has also demonstrated various potential applications in geologic exploration, compositional detection, process monitoring, etc. However, due to the limitation of the outer contour or surface morphology of the monitoring object, it is inevitable to use LIBS technology for exploration, detection, and monitoring. However, due to the limitations of the monitoring object's outer contour or surface morphology, when LIBS technology is used for detection, testing and monitoring, it is inevitable that the laser will act on the material surface in a non-perpendicular incidence mode, which will lead to fluctuations in the intensity and stability of the induced plasma spectra and affect the results of the geological detection, compositional testing and process monitoring based on LIBS technology. Therefore, in order to improve the accuracy of LIBS analysis, it is necessary to consider the effect of laser incidence angle on the intensity and repeatability of LIBS spectra. For the online monitoring of laser paint removal LIBS on free-form surfaces or geometrically mutated regions of aircraft skin, the paper collected LIBS spectra under different laser incidence angles, used baseline correction to preprocess the data of the original spectra, and selected six characteristic spectral lines as the analysis spectra, and studied the changing rules and causes of the laser incidence angle on the intensity of the characteristic signals of the LIBS spectra and their repeatability. The results show that in the range of laser incidence angle from 90° to 60° (5° interval), the intensity of the characteristic peaks decreases with the decrease of the laser incidence angle in a general pattern of “increasing first and then decreasing”, in which the characteristic spectral line Ti Ⅰ 429.926 nm has a better sensitivity and response to the laser incidence angle. When the laser incidence angle is 75°, the characteristic peak intensity of the characteristic spectral line Ti Ⅰ 429.926 nm increases by nearly 145% compared with that of the laser incidence angle of 90°，which indicates that the selection of the laser incidence angle is conducive to the enhancement of the spectral signal quality and signal intensity. In addition, the characteristic peak intensity of the characteristic spectral line Ti Ⅰ 429.926 nm has a better repeatability in the range of 75°～65° laser incidence angle, while the characteristic peak intensities of the rest of the elements have a better repeatability at the laser incidence angle of 85°. This study illustrates the connection and difference of the LIBS feature spectra when the laser is incident in vertical and non-vertical ways, and reveals the mechanism of the laser incidence angle on the multi-dimensional characteristics of the LIBS feature signals, which can be used as a reference for the detection, testing, and monitoring of LIBS when the laser is not incident vertically.
		2025 Vol. 45 (07): 1999-2007 [Abstract] ( 8 ) PDF (21457 KB) ( 3 )

	2008	Universal High Fidelity Spectral Image Compression Based on Color Perception
		LIANG Wei^{1, 2}, CAI Lei^{1, 2}, HAO Wen^{1, 2}, JIN Hai-yan^{1, 2}, HOU Yu³
		DOI: 10.3964/j.issn.1000-0593(2025)07-2008-09
		Aiming at the application of spectral images in the fields of color high fidelity reproduction in specific reproduction environments, this paper proposes universal low-complexity, color-high-fidelity spectral image compression methods based on visual characteristics in specific illumination, which could enhance algorithms' versatility, improve compression efficiency, and further facilitate images' storage and transmission. This paper first studies the color reproduction principle of spectral images in specific reproduction environments, designs a measurement method for the color error of reconstructed spectral images, and then proposes distortion guidelines for spectral image color fidelity compression in specific illumination. Based on the color distortion guidelines, the compression principle is derived. Then spectral preprocessing, spatial-spectral de-redundancy methods, encoding methods, and optimization strategies are designed, and finally, spectral image compression methods for high-fidelity reproduction are proposed. In terms of distortion guidance criteria, first the color decomposition environment of spectral images is constructed, and a matrix operator is proposed to extract color perception information from spectral images under specific lighting (single or mixed lighting); then, through the color perception information extraction operator, color perception error is used to measure the deviation of spectral images in the color measurement; finally, the color perception distortion criterion of the spectral image is proposed to guide the compression process. Based on this criterion, a targeted compression principle is proposed, and the compression flow of this paper is designed. First, the color perception weighted preprocessing of the spectral data is performed, and the color perception information extraction operator is used to obtain spectral color perception data under specific reproduction conditions that still maintains spectral characteristics; then, based on the color perception compression principle, APWS-RA encoding is performed on the color perception spectral data. The method is named W_SF-APWS-RA. Spectral image decoding is divided into two stages. First, the compressed bit stream is encoded inversely to form a reconstructed spectral color perception data matrix; then, the reconstructed spectral image is obtained by multiplying the inverse matrix of the perception information extraction operator and the reconstructed spectral color perception data matrix. Experiments show that at the same bit rate, compared with the existing low-complexity APWS, APWS-RA, and universal color-high-fidelity WF-APWS-RA compression, WsF-APWS-RA codings can not only more effectively retain spectral color information under specific reproduction conditions, but also have the best variable illumination color reproduction stability. Meanwhile, it can also effectively improve the accuracy of spectral reconstruction. Therefore, the new methods can also be generalized to remote sensing and other fields, and have important practical value.
		2025 Vol. 45 (07): 2008-2016 [Abstract] ( 10 ) PDF (41210 KB) ( 5 )

	2017	Prediction of Soil Total Nitrogen Based on NIR Spectroscopy and BiGRU Model With Attention Mechanism
		JU Wei-liang¹, YANG Wei^{1, 2}, SONG Ya-mei¹, LIU Nan¹, LI Hao¹, LI Min-zan^{1, 2*}
		DOI: 10.3964/j.issn.1000-0593(2025)07-2017-09
		Soil total nitrogen (STN) content is a key indicator for evaluating soil fertility, and its accurate measurement is of great significance for improving crop yield and quality. Predicting soil total nitrogen content using near-infrared spectroscopy has been proven to be an effective solution. However, due to the high dimensionality and complex time series characteristics of soil spectral data, traditional models often struggle to capture critical information, affecting prediction accuracy. Hence, based on near-infrared spectra (900～1 700 nm) of 600 soil samples, the method for predicting soil total nitrogen (STN) content was investigated,and a Bidirectional Gated Recurrent Unit based on an Attention Mechanism (BiGRU-Attention) model was proposed. First, the spectral data quality was optimized using SG filtering and SNV preprocessing methods. Then, using the CARS feature selection algorithm, the wavelengths for modeling were reduced from 198 to 30, thereby removing redundant information and decreasing the complexity of the modeling process. The BiGRU-Attention model effectively manages the flow of information using update and reset gates and enables the model to disregard the unimportant spectral data and retain the key information, which impacts the prediction accuracy. By leveraging the dual temporal sequence processing advantages of bidirectional GRU, the model can simultaneously handle forward and backward inputs of spectral sequences, thereby enhancing its ability to focus on edge data and comprehensively capture the dependencies present in soil spectral data. Additionally, the model employs an attention layer to compute the importance of each segment using the QKV matrix and dynamically determines which features should be emphasized based on the sequential interdependencies. This process calculates attention weight matrices to assign weights to each input data point, generating a more relevant context matrix that improves the model's predictive accuracy. Experimental results show that the BiGRU-Attention model can better understand the correlation between bands and perform better in prediction than other models, with the spectral data achieving an R² of 0.87 and an RMSE of 0.20 g·kg^-1 on the test dataset after feature selection. This study provides technical support for rapid soil nutrient detection and offers a method and reference for establishing high-accuracy STN prediction models.
		2025 Vol. 45 (07): 2017-2025 [Abstract] ( 5 ) PDF (8185 KB) ( 3 )

	2026	Monte Carlo-Based Full Transmission Light Transmission Simulation of Multi-Tissue Layers of Pomelo Fruit and Non-Destructive Testing of Internal Quality
		CHEN Xin¹, XU Sai^2*, LU Hua-zhong³, LIANG Xin²
		DOI: 10.3964/j.issn.1000-0593(2025)07-2026-08
		The thick peel and large volume of pomelos result in low spectral signal intensity, leading to poor performance in nondestructive internal quality detection. This study proposes a Monte Carlo-based simulation and experimental verification for the full-transmission light transport and nondestructive internal quality detection of pomelos with multiple tissue layers. By measuring the optical properties of the multiple tissue layers of pomelos, we simulated the photon travel distance and interaction probabilities within these layers. The light transport results were obtained after simulating the photon interactions with the multilayer tissues. Following this, modeling simulations were conducted based on the optical properties of these layers. We varied the incident angle of the light source and the rotation angle of the detector to find the optimal angles for light transmission. Finally, an experimental platform was constructed to verify the findings. The near-infrared spectral data underwent preprocessing steps such as Savitzky-Golay (SG) smoothing, standard normal variate (SNV) transformation, and competitive adaptive reweighted sampling (CARS) for feature extraction, followed by partial least squares regression (PLSR) modeling. The study results indicated that photons traveled the longest distance and experienced the greatest attenuation with the lowest survival probability in the epicarp oil cell layer, whereas photons traveled the shortest distance and had the highest survival probability in the pulp lobes. The optimal parameters for light transmission were a light source incident angle of 36° and a detector rotation angle of 10° around the Z-axis. The modeling results with these optimal parameters showed R² and RMSEC values of 0.89 and 0.25 for the training set, and R² and RMSEP values of 0.84 and 0.38 for the prediction set. In contrast, without parameter optimization, the results showed a light source incident angle of 0° and a detector rotation of 0°, with the training set R² and RMSEC being 0.85 and 0.27, and the prediction set R² and RMSEP being 0.80 and 0.34. For a light source incident angle of 18° and a detector rotation of 10° around the Z-axis, the training set R² and RMSEC were 0.80 and 0.34, while the prediction set R² and RMSEP were 0.73 and 0.74. For a light source incident angle of 36° and a detector rotation of 10° around the Y-axis, the training set R² and RMSEC were 0.69 and 0.25, while the prediction set R² and RMSEP were 0.60 and 0.83. The findings of this study on the light transport parameters for multiple tissue layers of pomelos can improve the effectiveness of nondestructive internal quality detection. Additionally, these results provide a reference for the simulation and experimental nondestructive internal quality detection of other multi-tissue layer fruits using full-transmission light transport.
		2025 Vol. 45 (07): 2026-2033 [Abstract] ( 6 ) PDF (9563 KB) ( 7 )

	2034	Infrared and Visible Image Fusion Based on Improved Latent Low-Rank and Unsharp Masks
		FENG Zhun-ruo¹, LI Yun-hong^1*, CHEN Wei-zhong¹, SU Xue-ping¹, CHEN Jin-ni¹, LI Jia-peng¹, LIU Huan¹, LI Shi-bo²
		DOI: 10.3964/j.issn.1000-0593(2025)07-2034-11
		To address the challenges of incomplete salient information extraction and detail degradation in infrared and visible light image fusion under low-light conditions, we propose an enhanced fusion algorithm that integrates Latent Low-Rank Representation (LatLRR) with Anisotropic Diffusion-Based Unsharp Mask(ADUSM). Initially, we apply block-wise segmentation and vectorization to the infrared and visible images, subsequently inputting them into the LatLRR model. Through an inverse reconstruction operation, we extract low-rank components from the infrared images and obtain basic salient components from the visible images. Next, the basic salient components undergo processing with ADUSM for pixel-wise differencing, allowing for further decomposition into deep salient detail components and multi-level detail features. Subsequently, the low-rank components are fused utilizing a visual saliency map rule, which enhances the retention and visibility of salient targets in the resultant fused image. For the deep salient detail components, we employ local entropy maximization for fusion, establishing a maximum activity coefficient to preserve the deep salient details effectively, thereby improving the overall quality and visual richness of the fused image. The multi-level detail features are fused using a weighted average strategy based on maximum spatial frequency, which adapts to the multi-level detail features of the input images, thus enhancing the overall clarity and contrast. Finally, we conduct a comparative analysis of our proposed method against Bayesian, Wavelet, LatLRR, MSVD, and MDLatLRR algorithms using the TNO and M3FD datasets. Experimental results demonstrate that our algorithm significantly outperforms traditional low-rank algorithms in average gradient methods, achieving enhancements of 31%, 2.1%, 4.4%, and 34% in average gradient, information entropy, standard deviation, and spatial frequency metrics. Comprehensive subjective and objective evaluations indicate that the fused images produced by our method not only exhibit rich texture details and clear salient targets but also present substantial advantages over various competing methods. This study effectively addresses the issue of incomplete salient information extraction in low-light environments, exhibiting robust generalization capabilities. The integration of improved Latent Low-Rank and ADUSM filtering is demonstrated to be both effective and feasible in the realm of infrared and visible light image fusion, offering significant scientific contributions to the advancement and application of this technology.
		2025 Vol. 45 (07): 2034-2044 [Abstract] ( 5 ) PDF (42738 KB) ( 3 )

	2045	Hyperspectral Mural Image Inpaint Based on Spatial-Spectral Enhancement Transform
		ZHANG Mian¹, ZHAO Jia-yu^1*, ZHOU Han², LIAN Yu-sheng²
		DOI: 10.3964/j.issn.1000-0593(2025)07-2045-09
		The non-destructive murals inpainting and colored paintings are an important topic and research hotspot for protecting and inheriting architectural cultural heritage. Hyperspectral imaging technology can simultaneously obtain two-dimensional spatial information and one-dimensional spectral information of targets.It has become an important technical means for digital collection, restoration, and analysis of cultural relics to conduct spectral digital collection and non-destructive analysis of murals and painted cultural relics without contact and without independent samples. Existing RGB color mural inpainting methods can not realize the collection, restoration and analysis of multi band spectral information in hyperspectral images; In addition, the existing depth generating color mural inpainting methods based on convolutional neural network have shortcomings such as insufficient modeling ability of spatial structure and spectral characteristics, weak ability of global information exploration and modeling, which seriously affect the mural inpainting accuracy. In order to solve the above problems, this paper proposes a hyperspectral mural inpainting method based on the space spectrum enhancement Transformer. Firstly, the hyperspectral mural inpainting to be repaired is reduced in spectral dimension and converted into an RGB color image. Then, the space and color information of the RGB color image is repaired by using the proposed generation countermeasure network based on the space spectrum enhanced Transformer. The repair network proposed in this paper is divided into a spatial information pre-repair network (Spa-PIN) and a spatial color information repair network (Spa-Color-IN). The effective repair of mural images is achieved by combining a spatial attention and spectral attention module (SAESA). In the reconstruction phase of spatial information structure, the basic shape and texture reconstruction of mural inpainting are emphasized. In the phase of color restoration, spatial attention and spectral attention are enhanced to improve the quality of restoration. Finally, using the proposed clustering BPNN, the dimension of the repaired RGB image is upgraded and reconstructed, and the repaired target hyperspectral image data cube is obtained. The attention mechanism of the space spectrum enhancement Transformer proposed in this paper performs spatial coordinate convolution fusion and spectral cube local global attention fusion on image features, which can simultaneously model the spatial spectrum correlation of the image in the global and local ranges, and enhance the repair ability of spatial spectrum details. The experimental results on the public datasets show that, compared with the current three advanced restoration methods, the method proposed in this paper has the optimal quantitative indicators and mural inpainting effect. It can effectively and accurately restore hyperspectral mural inpainting and provide new advanced technical means for high-precision collection, restoration, and analysis of architectural heritage such as mural inpaintings.
		2025 Vol. 45 (07): 2045-2053 [Abstract] ( 6 ) PDF (28217 KB) ( 3 )

	2054	Analysis and Correction of On-Orbit Spectral Drift of FY-3E Solar Spectral Irradiance Monitor Visible Band
		TAN Xiao-feng^{1, 2}, QI Jin^{2, 3, 4}, LI Zhan-feng⁵, ZHANG Peng^{6, 7}
		DOI: 10.3964/j.issn.1000-0593(2025)07-2054-07
		The Solar spectral irradiance monitor (SSIM) is China's first spaceborne solar irradiance spectrometer designed to measure solar continuum spectrumat the top of the atmosphere. It is mounted on the Fengyun-3E (FY-3E) satellite, the fifth satellite in the second generation of Chinese polar-orbit meteorological satellites. SSIM covers a spectral range from 165 to 2 400 nm, with a spectral resolution of 1 nm and sampling intervals of 0.1 and 0.25 nm in the ultraviolet and visible bands, which can capture fine solar spectrum characteristics of the sun. This study evaluated the spectral calibration accuracy of the SSIM visible band using the latest Total and Spectral Solar Irradiance Sensor-1 Hybrid Solar Reference Spectrum (TSIS-1 HSRS) as reference, with measurements from September 2021 to November 2023. By convoluting the pre-launch spectral slit function with TSIS-1 HSRS, a reference spectrum was generated to match the SSIM's characteristics. Using the spectrum matching method, nine solar Fraunhofer lines were selected to characterize wavelength shifts. The temporalvariationof spectral drift was analyzed, and a refined correction method was developed to improvethe spectral calibration accuracy of SSIM. Considering the characteristics of the early morning orbit, the annual variation of the Doppler shift was first analyzed, revealing the maximum wavelength shift of -0.005 to 0.001nm at 700 nm. Based on the assessment results of the Fraunhofer lines, the long-term variation of wavelength shifts in the visible band exhibited periodic fluctuations, with the shift of 410 nm varying from -0.032 to 0.025 nm compared to the initial on-orbit state. Further analysis indicates that the spectral driftis strongly correlated with the periodic fluctuations in the grating temperature of SSIM, with correlation coefficients between the wavelength shifts of the Fraunhofer lines and the grating temperature ranging from -0.766 to -0.964. Considering the drift characteristics on different spectral regions, the method of segmenting and sliding to calculate wavelength shifts and then fitting them for correction has been applied to the SSIM visible band, which shows that the results of spectral calibration accuracy are better than 0.028 nm. This method eliminates the impact of satellite launch, environmental changes on spectral accuracy, and also temperature fluctuations induced spectral drift. It has good long-term stability and effectively improves the on-orbit spectral calibration accuracy of SSIM. This paper provides a reference for the research on the on-orbit spectral calibration of wide-band solar spectrometers.
		2025 Vol. 45 (07): 2054-2060 [Abstract] ( 9 ) PDF (4820 KB) ( 7 )

	2061	Rapid Inversion of Gold Ore Grades Based on Hyperspectral Data and Stacking Ensemble Learning Algorithm
		MAO Ya-chun, XIA An-ni^*, CAO Wang, LIU Jing, WEN Jie, HE Li-ming, CHEN Xuan-he
		DOI: 10.3964/j.issn.1000-0593(2025)07-2061-07
		The gold mining resource holds significant economic and financial value, providing precious metal resources for the country, driving economic growth, and enhancing currency stability and hedging capabilities in the international financial market. However, while precise, the current chemical analysis methods for measuring gold ore grades in mines face issues such as long processing times, high costs, and reagent pollution, hindering the automation of ore grade and beneficiation method adjustments based on real-time grade information. In contrast, due to its efficiency, eco-friendliness, and in-situ measurement advantages, visible-near infrared spectroscopy is gradually becoming an effective alternative for estimating metal grades in mining areas. First, the raw spectral data were processed using Savitzky-Golay (SG) smoothing to reduce noise, and the spectral characteristics of gold ores were analyzed. It was found that reflectance correlates with gold grade, and a gold absorption feature is present at 455 nm. Based on this finding, dimensionality reduction was performed on the raw spectral data using principal component analysis (PCA), isometric feature mapping (ISOMAP), and locally linear embedding (LLE), with the resulting dimensions reduced to 6, 5, and 5, respectively. Finally, prediction models for gold grade were established using random forest (RF), extremely randomized trees (ET), decision trees (DT), gradient boosting decision tree (GBDT), adaptive boosting (Adaboost), extreme gradient boosting (XGBoost), and stacking ensemble learning algorithms on the dimensionally reduced data.Results indicated that the Stacking ensemble learning method outperformed single models in all aspects. Among them, the LLE-Stacking combined model achieved the highest accuracy, withR² of 0.972, RPD of 5.935, and an average relative error of 0.231 between predicted and actual values. The method proposed in this study allows for rapid and accurate predictions of gold content in ore, significantly improving the inversion accuracy compared to traditional models, providing new technological means for the rapid and in-situ measurement of gold grades in mines, and holding great significance for efficient gold extraction.
		2025 Vol. 45 (07): 2061-2067 [Abstract] ( 5 ) PDF (22795 KB) ( 3 )

	2068	Estimation Approach of Chlorophyll-a Concentration in Baiyangdian Based on Semi-Supervised Optical Classification
		CHEN Wen-yue^{1, 2}, ZHAO Qi-chao^{1, 2*}, YANG Xiu-feng^{1, 2}, HAN Bao-hui^{1, 2, 3}, ZHANG Yu-qing^{1, 2}
		DOI: 10.3964/j.issn.1000-0593(2025)07-2068-10
		The estimation of Chlorophyll-a concentration (Chl-a) using remote sensing technology is considered an effective way to monitor eutrophication in water bodies. Due to the complexity of the optical properties of inland water bodies and the existence of large spatial and temporal variability, it is difficult for a single estimation model to accurately estimate Chl-a concentration, and targeted modelling estimation based on the results of the optical classification of water bodies is one of the most important technological approaches for inland water body Chl-a inversion. In this study, Baiyangdian is taken as the study area, and a semi-supervised optical classification-based estimation method is proposed using the measured reflectance spectra and Chl-a concentration as the data source. First, to ensure that the number of samples in each category after classification is sufficient to support the construction of the estimation model, this study divides the samples into modelling set and validation set according to the ratio of 7∶1. Representative labelled samples of Baiyangdian were selected by multiple spectral indices and Moore's voting algorithm. Secondly, the fuzzy C-mean clustering algorithm and random forest algorithm are selected to construct a semi-supervised classifier, based on the representative labelled samples obtained, to further explore the potential information in the unlabelled samples and improve the classification accuracy of the unlabelled samples. Finally, the estimation models were established for each water type, the centre-of-mass spectra of each water type were calculated, and the final estimation results were obtained by hybrid-weighting using distance weights. The results show that Baiyangdian water bodies can be classified into phytoplankton-dominated, intermediate and suspended matter-dominated according to the spectral characteristics, and different types of water bodies have obvious differences in optical properties, which can be used to select the optimal estimation model and improve the estimation accuracy according to the optical classification results. Compared with the traditional optical classification strategy, the method proposed in this study performed the best, with a decrease in the mean relative error, root mean square error and mean absolute error, and was able to estimate Chl-a concentration more accurately (MRE=0.10, RMSE=0.126 μg·L^-1, MAE=0.106 μg·L^-1). In addition, applying ZY01-02E image data for Chl-a estimation in this study can effectively reveal the spatial distribution of Chl-a concentration. This method demonstrates the potential for application in eutrophication monitoring of water bodies, and also provides a new idea for remote sensing estimation of Chl-a concentration in inland water bodies.
		2025 Vol. 45 (07): 2068-2077 [Abstract] ( 6 ) PDF (12834 KB) ( 2 )

	2078	Research on the Rammed Earth Construction Materials of Shang Dynasty Capital Sites in Zhengzhou
		GU Tian-yang^{1, 2}, SHI Dong-hui³, YANG Shu-gang³, SONG Guo-ding^4, ZHANG Yu-xiu^5
		DOI: 10.3964/j.issn.1000-0593(2025)07-2078-09
		Rammed earth constructions are recognized as important archaeological discoveries of the Shang Dynasty in Zhengzhou, providing physical evidence for understanding the architectural materials and techniques of the period and containing a wealth of information related to social organization, production activities, and cultural phenomena. This study focused on the rammed earth at the Zhengzhou Shang City and Xiaoshuangqiao site, through the determination of hardness, porosity, moisture content limits, as well as X-ray diffraction analysis, X-ray fluorescence analysis, and scanning electron microscopy observations, a deeper understanding of the elemental composition, physical properties, and mechanical performance of rammed earth materials has been achieved. At the same time, an analysis of the compositional characteristics of the rammed earth materials, the changes in internal structure and properties before and after rammed construction, and the characteristics of human activities reflected within has been conducted. Based on the archaeological background information of the samples, it is believed that the rammed earth quality of the constructions at the Shang Dynasty capital cities in Zhengzhou is relatively high. During this stage, the rammed earth construction technique is considered to be in a period of stable development, with most of the raw materials being sourced from the common clay found near the sites. Significant differences in the composition and properties of the rammed earth samples from certain specific building types or parts have been observed, indicating that the choice and handling of rammed earth raw materials may be related to the functional requirements of different construction types and parts, based on the ancestors' preliminary understanding of the regional environment and soil characteristics. Reflections and prospects for analyzing rammed earth technology are also proposed based on summarized experiences, which provide valuable references for future research.
		2025 Vol. 45 (07): 2078-2086 [Abstract] ( 6 ) PDF (26945 KB) ( 4 )

	2087	Raman Measurement of Uptake Coefficient for Heterogeneous SO₂ Reactions Catalyzed by Transition Metal Ions on the Surface of Atmospheric Fine Particulate Matter
		CAO Xue¹, SUN Jiu-yi¹, WANG Cai-li², LI Ke-shu³, CAI Hua²
		DOI: 10.3964/j.issn.1000-0593(2025)07-2087-06
		Fine particulate matter (PM_2.5) pollution was mainly caused by micron and submicron aerosol particles or micro-droplets generated by human activities. Still, the formation mechanism remains unclear, and corresponding effective solutions are lacking. The precise prevention and control of urban PM_2.5 pollution in China depends on our in-depth understanding of the formation mechanism of secondary aerosols. The current bottleneck and difficulty lie in the fact that aerosols' key physicochemical parameters (such as uptake coefficients) and their precursors still cannot be accurately measured. This will greatly limit our understanding of the evolution laws, such as the formation, rapid growth, and collision recombination of secondary aerosols. Therefore, to understand the formation mechanism of atmospheric particulate matter, it is necessary to accurately measure the key physicochemical parameters in the evolution process of fine particulate matter. Sulfate was an important component of PM_2.5 in the atmosphere of China. However, the current chemical models underestimate the sulfate concentration in PM_2.5 haze pollution events. There were still many uncertainties regarding the surface oxidation process of SO₂ in aerosol droplets and the catalytic process of transition metal ions(TMI) in droplets, indicating that our understanding of the transformation mechanism of SO₂ in the atmosphere was still insufficient. Obtaining the kinetic data of sulfate secondary transformation, determining the rapid growth mechanism of secondary inorganic aerosols, and accurately determining the uptake reaction rate of trace gases in the atmosphere on the surface of particulate matter were the key parameters for quantitative analysis of atmospheric heterogeneous reactions. This study employed a laser confocal micro-Raman spectrometer to measure the kinetics of SO₂ oxidation catalyzed by Fe(Ⅲ) in ammonium chloride droplets. Based on the temporal variation of the spontaneous Raman signal of SO^2-₄, we established an accurate method for measuring the uptake coefficient of SO₂ in heterogeneous reactions with a single droplet. We investigated the reaction kinetics between trace SO₂ gas and a single droplet under external field conditions. This method validated the surface kinetic processes, and determined the reaction uptake coefficient.
		2025 Vol. 45 (07): 2087-2092 [Abstract] ( 5 ) PDF (4981 KB) ( 9 )

	2093	Fluorescence Spectral Characteristics of Dissolved Organic Matter in Typical Reservoirs in the Yellow River Basin
		WU Li¹, CHANG Miao¹, LAI Meng-yuan¹, ZHAO Tong-qian^1*, WANG Liang²
		DOI: 10.3964/j.issn.1000-0593(2025)07-2093-08
		Xiaolangdi Reservoir is the largest hydraulic engineering project in the Yellow River Basin, and the annual water-sediment regulation can reduce siltation in the downstream river. However, reservoir storage will change the hydrodynamic conditions of the river. Exploring the differences in the spectral characteristics of dissolved organic matter (DOM) before and after reservoir storage is crucial to reveal the changes in the aquatic ecosystem of the Yellow River Basin. In 2021, water samples were collected before, during, and after the water-sediment regulation period. The characteristics of the change in DOM fluorescence spectral parameters in the reservoir were analyzed. The DOM components and their sources were resolved using three-dimensional fluorescence spectroscopy coupled with the parallel factor method. Combined with the water quality parameters of the water body, the key factors affecting the changes in DOM fluorescence characteristics in the reservoir were elucidated by correlation analysis and redundancy analysis (RDA). Results indicated that (1) the fluorescence index and humification index of DOM did not change remarkably in all periods. The DOM of the reservoir water body was characterized by strong authigenicity and weak humification. During the water-sediment regulation, the authigenic index and freshness index of DOM decreased remarkably, and the newly authigenic DOM produced by microbial activities and other activities decreased in the reservoir area. (2) Four fluorescent components were resolved in the DOM of the Xiaolangdi Reservoir. Differences in the composition of the components were observed in different periods. The C1 and C2 components were humus-like substances, and the C3 and C4 components were protein-like in the non-water-sediment regulation period. On the contrary, the C1, C2, and C3 components were humus-like substances, and the C4 component was a protein-like substance in the water-sediment regulation period. Reservoir DOM before water-sediment regulation was dominated by protein-like substances (52.95%). The exogenous DOM entered the water body because of reservoir drainage and sediment discharge during the regulation period, which led to a remarkable increase in the proportion of humus-like substances (78.53%). The proportion of humus-like substances after water-sediment regulation was 58.80%. (3) Correlation analysis showed that water temperature (WT), flow, and sand content were the main factors influencing the source of DOM in Xiaolangdi Reservoir. During the non-water-sediment regulation period, DOM concentration was mainly affected by the exogenous input of humus-like substances. The degradation of tyrosine components may be related to pH and electrical conductivity(EC). RDA found that pH, WT, and EC were the main environmental factors affecting the differences in DOM fractions in different periods. The results of this study can provide basic data for the study of the fluorescence characteristics of DOM in the Yellow River Basin reservoirs, as well as the scientific basis for further revealing the influence of water body disturbance on the carbon cycle in the Yellow River Basin.
		2025 Vol. 45 (07): 2093-2100 [Abstract] ( 7 ) PDF (11150 KB) ( 3 )

　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