[1]安婧,韩俊英.基于高光谱成像的苹果品质指标的无损检测方法[J].江苏农业学报,2026,42(05):973-981.[doi:doi:10.3969/j.issn.1000-4440.2026.05.011]
　AN Jing,HAN Junying.A non-destructive detection method for apple quality indices based on hyperspectral imaging[J].,2026,42(05):973-981.[doi:doi:10.3969/j.issn.1000-4440.2026.05.011]
点击复制

基于高光谱成像的苹果品质指标的无损检测方法()

江苏农业学报[ISSN:1006-6977/CN:61-1281/TN]

卷:
42
期数:
2026年05期
页码:
973-981
栏目:
农业信息工程
出版日期:
2026-05-31

文章信息/Info

Title:
A non-destructive detection method for apple quality indices based on hyperspectral imaging
作者:
安婧; 韩俊英
(甘肃农业大学信息科学技术学院,甘肃兰州730070)
Author(s):
AN Jing; HAN Junying
(College of Information Science and Technology, Gansu Agricultural University, Lanzhou 730070, China)
关键词:
高光谱成像; 苹果品质检测; 机器学习; 图像处理
Keywords:
hyperspectral imaging; apple quality detection; machine learning; image processing
分类号:
TS255.7;O657.3
DOI:
doi:10.3969/j.issn.1000-4440.2026.05.011
文献标志码:
A
摘要:
为实现苹果内部品质的快速无损检测,本研究基于380～1 018 nm高光谱成像技术构建了融合多种预处理、特征选择与建模算法的预测框架。采用Savitzky-Golay(SG)平滑、离散小波变换(DWT)、多元散射校正(MSC)、标准正态变量变换(SNV)、自适应加权平滑(AWS)、鲁棒性主成分分析(RPCA)和基线校正(BC)预处理光谱数据,结合连续投影算法(SPA)、竞争性自适应重加权采样法(CARS)和基于Bootstrap的优化光谱选择方法(BOSS)提取特征波段,构建偏最小二乘回归模型、支持向量回归模型与岭回归模型。结果表明,不同指标的最优建模路径呈现显著差异。对于维生素C含量,SG-SPA-SVR组合建模预测效果最佳。对于可滴定酸含量,BC-原始光谱-Ridge组合建模效果最佳。对于淀粉含量,BC-原始光谱-SVR组合建模预测效果最佳。对于可溶性固形物含量,MSC-原始光谱-Ridge组合建模预测效果最佳。本研究构建的多策略建模体系可为苹果品质无损检测及便携式高光谱系统开发提供理论支持与技术参考。
Abstract:
To achieve rapid and non-destructive detection of internal quality in apples, this study established a prediction framework integrating multiple preprocessing, feature selection, and modeling algorithms based on hyperspectral imaging technology in the range of 380-1 018 nm. Spectral data were preprocessed using Savitzky-Golay (SG) smoothing, discrete wavelet transform (DWT), multiplicative scatter correction (MSC), standard normal variate transformation (SNV), adaptive weighted smoothing (AWS), robust principal component analysis (RPCA), and baseline correction (BC). Characteristic wavelengths were selected using the successive projections algorithm (SPA), competitive adaptive reweighted sampling (CARS), and Bootstrap-based spectral selection (BOSS). Partial least squares regression, support vector regression, and ridge regression models were constructed. The results showed that the optimal modeling paths for different quality indices differed significantly. For vitamin C content, the combination of SG-SPA-SVR achieved the best prediction performance. For titratable acid content, the BC-raw spectrum-Ridge combination performed best. For starch content, the BC-raw spectrum-SVR combination showed the optimal prediction effect. For soluble solids content, the MSC-raw spectrum-Ridge combination yielded the best prediction results. The multi-strategy modeling system established in this study can provide theoretical support and technical reference for non-destructive detection of apple quality and the development of portable hyperspectral systems.

参考文献/References:

[1]高冲. 基于高光谱与深度学习的苹果品质检测方法及装置研究[D]. 泰安:山东农业大学,2022.
[2]BAI S H, TOOTOONCHY M, KMPER W, et al. Predicting carbohydrate concentrations in avocado and macadamia leaves using hyperspectral imaging with partial least squares regressions and artificial neural networks[J]. Remote Sensing,2024,16(18):3389.
[3]WANG X J, HAN J Y, LIU C Z, et al. Non-destructive assessment of apple internal quality using rotational hyperspectral imaging[J]. Frontiers in Plant Science,2024,15:1432120.
[4]ZHU D Y, HAN J Y, LIU C Z, et al. Modeling of flaxseed protein,oil content,linoleic acid,and lignan content prediction based on hyperspectral imaging[J]. Frontiers in Plant Science,2024,15:1344143.
[5]曾柯宜,刘禹彤,张倩,等. 基于多模态融合的玉米种子成熟度的无损检测[J]. 食品安全质量检测学报,2025,16(2):171-177.
[6]GAO W J, CHENG X, LIU X H, et al. Apple firmness detection method based on hyperspectral technology[J]. Food Control,2024,166:110690.
[7]WANG F, ZHAO C J, YANG H, et al. Non-destructive and in-site estimation of apple quality and maturity by hyperspectral imaging[J]. Computers and Electronics in Agriculture,2022,195:106843.
[8]QIAO M M, CUI T, XIA G Y, et al. Integration of spectral and image features of hyperspectral imaging for quantitative determination of protein and starch contents in maize kernels[J]. Computers and Electronics in Agriculture,2024,218:108718.
[9]肖宏辉,邓浩,李春霖,等. 基于高光谱成像技术的橙子可溶性固形物和可滴定酸含量无损预测研究[J]. 核农学报,2025,39(5):1012-1020.
[10]DONG J L, GUO W C. Nondestructive determination of apple internal qualities using near-infrared hyperspectral reflectance imaging[J]. Food Analytical Methods,2015,8:2635-2646.
[11]王宏恩,冯国红,徐华东,等. 基于可见-近红外光谱和深度森林的蓝莓成熟度判别[J]. 光谱学与光谱分析,2024,44(11):3280-3286.
[12]MO C, KIM M S, KIM G, et al. Spatial assessment of soluble solid contents on apple slices using hyperspectral imaging[J]. Biosystems Engineering,2017,159:10-21.
[13]ZHANG P P, WANG H W, JI H W, et al. Hyperspectral imaging-based early damage degree representation of apple:a method of correlation coefficient[J]. Postharvest Biology and Technology,2023,199:112309.
[14]XU D H, WANG H W, JI H W, et al. Quantitative evaluation of impact damage to apple by hyperspectral imaging and mechanical parameters[J]. Food Analytical Methods,2019,12:371-380.
[15]JIANG D Y, CHANG Q R, ZHANG Z J, et al. Monitoring the degree of mosaic disease in apple leaves using hyperspectral images[J]. Remote Sensing,2023,15(10):2504.
[16]ZHANG P, SHEN B B, JI H W, et al. Nondestructive prediction of mechanical parameters to apple using hyperspectral imaging by support vector machine[J]. Food Analytical Methods,2022,15(5):1397-1406.
[17]李泽祺,杨铮,周壮飞,等. 高光谱与宽度学习检测草莓叶片含水率[J]. 光谱学与光谱分析,2025,45(6):1534-1542.
[18]LIU Y F, ZHANG Y, JIANG D Y, et al. Quantitative assessment of apple mosaic disease severity based on hyperspectral images and chlorophyll content[J]. Remote Sensing,2023,15(8):2202.
[19]WANG Z J, WU S S, ZUO C Z, et al. Exploring the variability and heterogeneity of apple firmness using visible and near-infrared hyperspectral imaging[J]. LWT-Food Science and Technology,2024,192:115704.
[20]ZHAN C Y, MAO H Y, FAN R S, et al. Detection of apple sucrose concentration based on fluorescence hyperspectral image system and machine learning[J]. Foods,2024,13(22):3547.
[21] 张永亮,汪泓,肖玖军,等. 基于高光谱的山区耕地土壤有机质含量估测[J]. 江苏农业学报,2024,40(1):112-120.
[22]李贝贝,王伟宁,赵犇,等. 云烟87叶片生理指标对镁浓度的响应特征及高光谱反演[J]. 南方农业学报,2025,56(11):3341-3351.
[23]余德炤,江晓东,杨莹颖,等. 基于RGB与高光谱图像的小麦低温胁迫评估模型[J]. 南方农业学报,2025,56(1):97-110.
[24]李伟,宋睿,刘明江,等. 光学遥感在区域灌溉监测中应用的研究进展[J]. 排灌机械工程学报,2024,42(11):1157-1165.
[25]刘江凡,赵泽艺,李朝阳,等. 基于无人机多光谱遥感的苹果树冠层SPAD反演[J]. 排灌机械工程学报,2024,42(5):525-531.
[26]王宇,汪泓,肖玖军,等. 基于MCC-GAPLS-PLSR的辣椒叶绿素含量高光谱定量反演[J]. 江苏农业学报,2024,40(5):865-873.

相似文献/References:

[1]宋镇,姬长英,张波.基于高光谱技术融合图像信息的杏鲍菇干燥过程中含水率检测[J].江苏农业学报,2019,(02):436.[doi:doi:10.3969/j.issn.1000-4440.2019.02.026]
　SONG Zhen,JI Chang-ying,ZHANG Bo.Visualized determination of moisture content in Pleurotus eryngii during drying process based on hyperspectral imaging technology[J].,2019,(05):436.[doi:doi:10.3969/j.issn.1000-4440.2019.02.026]
[2]沈广辉,曹瑶瑶,刘馨,等.近红外高光谱成像结合特征波长筛选识别小麦赤霉病瘪粒[J].江苏农业学报,2021,(02):509.[doi:doi:10.3969/j.issn.1000-4440.2021.02.029]
　SHEN Guang-hui,CAO Yao-yao,LIU Xin,et al.Identification of Fusarium damaged kernels using near infrared hyperspectral imaging and characteristic bands selection[J].,2021,(05):509.[doi:doi:10.3969/j.issn.1000-4440.2021.02.029]

备注/Memo

备注/Memo:
收稿日期：2025-07-19基金项目：国家自然科学基金项目(32360437)；甘肃省自然科学基金项目(25YFNA040)作者简介：安婧(2000-),女,甘肃天水人,硕士研究生,研究方向为农作物品质检测。(E-mail)2750214516@qq.com通讯作者：韩俊英,(E-mail)hanjy@gsau.edu.cn
更新日期/Last Update: 2026-06-17