Nondestructive Determinations of Texture and Quality of Preserved Egg Gel by Hyperspectral Method
CHEN Yuan-zhe1, WANG Qiao-hua1, 2*, TIAN Wen-qiang1, XU Bu-yun1, HU Jian-chao1
1. College of Engineering, Huazhong Agricultural University, Wuhan 430070, China
2. Key Laboratory of Agricultural Equipment in Mid-Lower Yangtze River，Ministry of Agriculture and Rural Affairs, Wuhan 430070, China
Abstract：As an important quality parameter, texture can significantly affect the gelatin quality of preserved eggs. There is no effectively rapid detection method at present. In this study, hyperspectral imaging technology was used to predict preserved eggs’ textural characteristics and classify different qualities. Hyperspectral data of high-quality eggs, qualified eggs and unqualified eggs were collected. The original spectra were transformed by single and combined transformation to analyze the correction between one-dimensional spectral and textural hardness and springiness. It was found that the spectral reflectance after CR-FD transformation was most correlated with the hardness and springiness of the gel texture, and the maximum values were 0.882, 0.86 5 at 683, 715 nm, respectively; The hardness and springiness were taken as disturbance factors to explore the optimal research area of the hardness and springiness, the results showed that: When hardness was used as disturbance factor, autocorrelation peaks existed at 476, 539, 647, 672, 728 and 851 nm, Spectral signals at 483, 572, 657, 739 and 826 nm were more sensitive to springiness value. Therefore, two sensitive bands, 476~851 and 483~826 nm, were finally selected as the study regions for gel hardness and springiness, respectively. Comparing five different variable selection methods(SPA, CARS, GA, PSO and UVE), it was found that PSO-PLSR model had the highest detection accuracy: theR2p and RMSEP for predicted hardness were 0.826 and 0.874 with an RPD of 2. TheR2p and RMSEP for predicted springiness were 0.886 and 0.402 with an RPD of 1.9. Three different classifiers were used to predict preserved eggs, and the classification accuracy of high-quality eggs, qualified eggs and unqualified eggs reached 97%, 92% and 100%, respectively. The accuracy and generalization ability of the PLS-DA model were better than the BP and RF model based on the confusion matrix and ROC curves of prediction results. In conclusion, the hyperspectral technique can be used to predict preserved eggs’ textural characteristics and classify different quality of preserved eggs.
陈远哲，王巧华，田文强，徐步云，胡建超. 出缸期皮蛋凝胶质构特性与品质的高光谱无损检测[J]. 光谱学与光谱分析, 2022, 42(06): 1985-1992.
CHEN Yuan-zhe, WANG Qiao-hua, TIAN Wen-qiang, XU Bu-yun, HU Jian-chao. Nondestructive Determinations of Texture and Quality of Preserved Egg Gel by Hyperspectral Method. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2022, 42(06): 1985-1992.
[1] YU Pei, WANG Xiu-jun, XU Wen, et al(于 沛, 王修俊, 徐 雯, 等). Food Science(食品科学), 2021，42(19)：65.
[2] XUAN Xi-long, WANG Zhi-wei, HUANG Ai-lan, et al(玄夕龙, 王志威, 黄爱兰, 等). Journal of Anhui Science and Technology University(安徽科技学院学报), 2017, 31(3): 38.
[3] Ai Minmin, Zhou Quan, Guo Shangguang, et al. Food Hydrocolloids, 2019, 94: 11.
[4] ZHAO Yan, XU Ming-sheng, YAO Yao, et al(赵 燕, 徐明生, 姚 瑶, 等). Journal of Chinese Institute of Food Science and Technology(中国食品学报), 2019, 19(6): 36.
[5] LIU Gui-shan, ZHANG Chong, FAN Nai-yun, et al(刘贵珊, 张 翀, 樊奈昀, 等). Spectroscopy and Spectral Analysis(光谱学与光谱分析), 2020, 40(8): 2558.
[6] TIAN You-wen, WU Wei, LU Shi-qian, et al(田有文, 吴 伟, 卢时铅, 等). Food Science(食品科学), 2021, 42(19): 260.
[7] LIU Long, FU Mei-zhang, WANG Shu-cai(刘 龙, 付美章, 王树才). Journal of Huazhong Agricultural University(华中农业大学学报), 2012, 31(4): 524.
[8] Li C H, Hsieh C H, Hung C H, et al. Foods, 2021, 10(2): 394.
[9] WANG Qiao-hua, MEI Lu, MA Mei-hu, et al(王巧华, 梅 璐, 马美湖, 等). Transactions of the Chinese Society of Agricultural Engineering(农业工程学报), 2019, 35(24): 314.
[10] ZHANG Xiao-min, ZHANG Yan-ning, JIANG Hai-yi, et al(张小敏, 张延宁, 姜海益, 等). Transactions of the Chinese Society for Agricultural Machinery(农业机械学报), 2021, 52(6): 232.
