Research on Non-Destructive Prediction Method of Tomato Quality
Indicators Using Hyperspectral Imaging and Extreme Learning
Machine
HUANG Lian-fei1, WU Sha1, HUANG Ren-shuai1, 2*
1. School of Food Science and Engineering, Guiyang University, Guiyang 550005, China
2. Engineering Research Center for Non-destructive Testing of Agricultural Products, Guiyang University, Guiyang 550005, China
Abstract：Hyperspectral imaging technology, with its unique advantages of acquiring continuous and abundant spectral information of samples and effectively reflecting internal compositional characteristics, provides a new technical pathway for the rapid and non-destructive detection of key physicochemical indicators of fruits and vegetables. Based on this, this study focuses on tomatoes as the research object, targeting their key physicochemical indicators—including Soluble Solids Content (SSC), lycopene content, and vitamin C content—to explore rapid and non-destructive prediction methods for these indicators. To enhance the accuracy and robustness of the prediction model, the hyperspectral data were first preprocessed using the Standard Normal Variate (SNV) transformation to mitigate the interference of scattering and spectral drift, followed by the selection of effective feature bands based on the Genetic Algorithm (GA). Further, Backpropagation Neural Network (BP) and Extreme Learning Machine (ELM) prediction models were respectively constructed to compare their performance differences in predicting various quality indicators. The results showed that ELM outperformed the BP model in predicting lycopene, SSC, and vitamin C. Compared to the BP model, the correlation Coefficient of the Prediction set (R2p) for ELM increased by 7.5%, 11.4%, and 9.8%, respectively; the Root Mean Square Error (RMSEP) decreased by 25.0%, 22.2%, and 10.4%, respectively; and the Relative Prediction Deviation (RPD) increased by 20.3%, 25.3%, and 28.0%, respectively. Notably, the RPD values of the ELM model in predicting lycopene, vitamin C, and SSC all exceeded 2.6, reaching a good level of prediction accuracy. This study offers a reliable and efficient technical solution for the rapid and non-destructive detection of key quality indicators in tomatoes. Also, it lays a methodological foundation for the application and promotion of non-destructive quality detection technologies for fruits and vegetables.
[1] Shao Y, Shi Y, Qin Y, et al. Food Chemistry, 2022, 386: 132864.
[2] Cakmak H. Food Quality and Shelf Life. Academic Press, 2019, 303.
[3] Wang B, Sun J, Xia L, et al. Food Reviews International, 2023, 39(2): 1043.
[4] Xiang Y, Chen Q, Su Z, et al. Frontiers in Plant Science, 2022, 13: 860656.
[5] DIWU Peng-yao, BIAN Xi-hui, WANG Zi-fang, et al(第五鹏瑶, 卞希慧, 王姿方, 等). Spectroscopy and Spectral Analysis(光谱学与光谱分析), 2019, 39(9): 2800.
[6] Kim E, Park J J, Lee G, et al. Food Chemistry: X, 2024, 23: 101763.
[7] Yun Y H, Li H D, Deng B C, et al. TrAC Trends in Analytical Chemistry, 2019, 113: 102.
[8] Dutta T, Dey S, Bhattacharyya S, et al. Expert Systems with Applications, 2021, 181: 115107.
[9] Peón J, Recondo C, Fernández S, et al. Remote Sensing, 2017, 9(12): 1211.
[10] Huang G B, Zhu Q Y, Siew C K. 2004 IEEE international Joint Conference on Neural Networks, 2004, 2: 985.
[11] Bian X H, Li S J, Fan M R, et al. Analytical Methods, 2016, 8(23): 4674.
[12] Cai Y, Liu X, Zhang Y, et al. Pattern Recognition Letters, 2018, 116: 101.
[13] Polder G, Van Der Heijden G W A M, Young I T. Transactions of the ASAE, 2002, 45(4): 1155.
[14] Duckena L, Alksnis R, Erdberga I, et al. Foods, 2023, 12(10): 1990.
[15] Rahman A, Kandpal L M, Lohumi S, et al. Applied Sciences, 2017, 7(1): 109.
[16] ZHANG Ruo-yu, RAO Xiu-qin, GAO Ying-wang, et al(张若宇, 饶秀勤, 高迎旺, 等). Transactions of the Chinese Society of Agricultural Engineering(农业工程学报), 2013, 29(23): 247.
[17] Cui X, Guan Z, Morgan K L, et al. Horticulturae, 2022, 8(12): 1204.
[18] Shao Y, Liu Y, Xuan G, et al. Infrared Physics & Technology, 2022, 127: 104403.
[19] Wu F, Xu F, Liu W, et al. Foods, 2023, 12(16): 3100.
[20] Samat A, Du P, Liu S, et al. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2014, 7(4): 1060.
[21] Villaseñor-Aguilar M J, Padilla-Medina J A, Botello-lvarez J E, et al. Applied Sciences, 2021, 11(19): 9332.
[22] Fang L, Wei K, Feng L, et al. Foods, 2020, 9(12): 1881.
[23] Gürbüz Colak N, Eken N T, Ülger M, et al. Plant Science, 2020, 292: 110393.
[24] Zhao J, Peng P, Wang J. Open Computer Science, 2023, 13(1): 20230102.
[25] Zhang X N, Yang J H, Zhao Y X. Fraction and Fractional, 2022, 6: 401.
[26] Wang Y, Xiong F, Zhang Y, et al. Food Chemistry, 2023, 404: 134503.
