沃柑可溶性固形物含量高光谱检测模型的建立与优化

doi:10.3964/j.issn.1000-0593(2025)02-0492-09

摘要
参考文献
相关文章 (15)

全文: PDF (14641 KB)
输出: BibTeX | EndNote (RIS)

摘要：针对沃柑品质快速检测需求，提出一种基于高光谱图像数据的沃柑可溶性固形物含量（SSC）检测方法，并分析随贮藏时间变化的SSC伪彩色分布图。分别获取307个整果样本和227个半果样本以及它们的SSC数据。比较标准正态变换（SNV）、多元散射校正（MSC）、Savitzky-Golay（SG）滤波、归一化（NM）、一阶导数（FD）、标准化（SD）和小波变换（WT）对偏最小二乘回归（PLSR）模型检测性能的影响以选择光谱预处理方法；比较PLSR、最小绝对值收缩与选择算子（LASSO）回归、支持向量机回归（SVR）、人工神经网络（ANN）、决策树（DT）、随机森林（RF）、轻量级梯度提升机（LightGBM）模型对独立验证集的检测能力以确定最佳模型建立方法，并利用遗传算法（GA）筛选特征波长以优化模型。结果表明：采用FD预处理结合LASSO回归算法所建模型对整果SSC检测效果最优，验证集决定系数（R²_p）和验证集均方根误差（RMSEP）分别为0.925 7和0.976 5；SD预处理结合RF模型对半果SSC检测效果最好，其R²_p和RMSEP分别为0.896 3和1.063 0；GA能够滤除53.85%和50.58%的整果和半果波长变量数，基于选择变量的整果和半果最优建模算法分别为SVR和RF，其模型R²_p、RMSEP分别为0.918 9、1.017 3和0.895 3、1.084 3。研究结果为沃柑SSC高通量无损检测提供了一种可行方案。

关键词：沃柑；高光谱；可溶性固形物含量；无损检测；集成学习

Abstract：To develop a rapid measurement method of SSC in Fortunella margarita, the detection models based on hyperspectral imaging data were established and optimized by employing various preprocess and regression algorithms, and the pseudo-color distribution of SSC with storage time was analyzed. The 307 whole citrus and 227 hemisected citrus samples were involved in hyperspectral data collection and the SSC values. The effects of preprocessing, including standard normal variate (SNV), multiplicative scatter correction (MSC), Savitzky-Golay (SG) filtering, normalization (NM), first derivative (FD), standardization (SD), and wavelet transformation (WT), on the performance of the partial least squares regression (PLSR) model were compared to select the appropriate preprocessing method. Then, the detection models were established by using PLSR, least absolute shrinkage and selection operator (LASSO) regression, support vector machine regression (SVR), artificial neural networks (ANN), decision trees (DT), random forest (RF) and light gradient boosting machine (Light GBM) algorithms. Furthermore, the models were optimized using genetic algorithms (GA) to select characteristic spectral wavelengths. The results indicated that for the whole citrus samples, the FD preprocessing could extract more features, and the LASSO regression model performed better than other models with 0.925 7 and 0.976 5 as the prediction determination coefficient (R²_p) and root mean square error of prediction (RMSEP), respectively. For the hemisected samples, the RF model based on the spectral after SD preprocessing had higher R²_p at 0.896 3 and lower RMSEP at 1.063 0. The GA could remove 53.85% and 50.58% wavelength variables to reduce the computational complexity for the whole and hemisected sample spectral, of which the SVR model has R²_p at 0.918 9. RMSEP at 1.017 3 RF model having R²_p at 0.895 3 and RMSEP at 1.084 3 performed better than other models. The results provided a feasible solution for high-throughput, non-destructive detection of SSC of Fortunella margarita.

Key words：Fortunella margarita; Hyperspectral image; Soluble solids content; Non-destructive detection; Ensemble learning

收稿日期: 2024-02-03 修订日期: 2024-08-29

中图分类号:

S24

基金资助: 国家自然科学基金青年科学基金项目（61205153）资助

通讯作者: 刘洁 E-mail: liujie@mail.hzau.edu.cn

作者简介: 李炜琪，女，1999年生，华中农业大学工学院硕士研究生 e-mail: 1939346545@qq.com

引用本文:

李炜琪，王一帆，俞越，刘洁. 沃柑可溶性固形物含量高光谱检测模型的建立与优化[J]. 光谱学与光谱分析, 2025, 45(02): 492-500.
LI Wei-qi, WANG Yi-fan, YU Yue, LIU Jie. Establishment and Optimization of the Hyperspectral Detection Model for Soluble Solids Content in Fortunella Margarita. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2025, 45(02): 492-500.

链接本文:

https://www.gpxygpfx.com/CN/10.3964/j.issn.1000-0593(2025)02-0492-09 或 https://www.gpxygpfx.com/CN/Y2025/V45/I02/492

[1] DONG Jian-mei，LI Jing，LAI Xin-pu，et al（董建梅，李晶，赖新朴，等）. Non-wood Forest Research（经济林研究），2023，41（4）：142.
[2] GAO Li，LONG Yong，LI Hong-zhang，et al（高丽，龙勇，李红章，等）. Food and Fermentation Industries（食品与发酵工业），2023，49（22）：140.
[3] ZHANG Xin-xin，LI Pao，YU Mei，et al（张欣欣，李跑，余梅，等）. Food Science（食品科学），2022，43（1）：260.
[4] GUO Zhi-ming，ZHAO Chun-jiang，HUANG Wen-qian，et al（郭志明，赵春江，黄文倩，等）. Transactions of the Chinese Society for Agricultural Machinery（农业机械学报），2015，46（7）：227.
[5] Jia B B, Wang W, Ni X Z, et al. Chemometrics and Intelligent Laboratory Systems, 2020, 198: 103936.
[6] Ma T, Li X Z, Inagaki T, et al. Journal of Food Engineering, 2018, 224: 53.
[7] Xuan G T, Gao C, Shao Y Y. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 2022, 272: 121016.
[8] Ma T, Xia Y, Inagaki T, et al. Postharvest Biology and Technology, 2021, 174: 111440.
[9] Teerachaichayut S, Ho H T. Postharvest Biology and Technology, 2017, 133: 20.
[10] Zhang H L, Zhan B H, Pan F, et al. Postharvest Biology and Technology, 2020, 163: 111148.
[11] Li S Y, Song Q M, Liu Y J, et al. Postharvest Biology and Technology, 2023, 204: 112454.
[12] GAO Sheng，XU Jian-hua（高升，徐建华）. Food Science（食品科学），2023，44（2）：327.
[13] ZHANG Chen，ZHU Yu-jie，FENG Guo-hong（张晨, 朱玉杰, 冯国红）. Food and Fermentation Industries（食品与发酵工业），2023，49（18）：306.
[14] JIANG Xiao-gang，YAO Jin-liang，ZHU Ming-wang，et al（姜小刚，姚金良，朱明旺，等）. Laser Technology（激光技术），2023，47（5）：666.
[15] YANG Bao-hua，GAO Zhi-wei，QI Lin，et al（杨宝华，高志伟，齐麟，等）. Spectroscopy and Spectral Analysis（光谱学与光谱分析），2021，41（11）：3559.
[16] WANG Kai-yi，YANG Sheng，GUO Cai-yun，et al（王恺怡，杨盛，郭彩云，等）. Journal of Instrumental Analysis（分析测试学报），2022，41（3）：398.
[17] ZHANG Chi，GUO Yuan，LI Ming（张驰，郭媛，黎明）. Computer Engineering and Applications（计算机工程与应用），2021，57（11）：57.
[18] WANG Dao-ming，LU Chang-hua，JIANG Wei-wei，et al（王道明，鲁昌华，蒋薇薇，等）. Journal of Electronic Measurement and Instrumentation（电子测量与仪器学报），2015，29（4）：611.
[19] Cai W T, Wei R H, Xu L H, et al. Information Processing in Agriculture, 2022, 9(3): 343.
[20] Cayuela J A. Postharvest Biology and Technology, 2008, 47(1): 75.
[21] Zerbini P E, Vanoli M, Rizzolo A, et al. Postharvest Biology and Technology, 2015, 101: 58.
[22] Zheng Y J, Tian S J, Xie L J. Postharvest Biology and Technology, 2023, 200: 112313.
[23] Wang A C, Hu D, Xie L J. Journal of Food Engineering, 2014, 126: 126.