食品与发酵工业 >

2025 , Vol. 51 >Issue 2: 311 - 319

DOI: https://doi.org/10.13995/j.cnki.11-1802/ts.039523

分析与检测

基于高光谱特征融合的榛子霉变检测方法研究

  • 张冬妍 ,
  • 毛思雨 ,
  • 杨子健 ,
  • 陈诺 ,
  • 吴晨旭 ,
  • 马苗源
展开
  • (东北林业大学 计算机与控制工程学院,黑龙江 哈尔滨,150040)
第一作者:博士,副教授(通信作者,E-mail:nefuzdhzdy@nefu.edu.cn)

收稿日期: 2024-04-11

  修回日期: 2024-06-11

  网络出版日期: 2025-02-14

基金资助

中央高校基本科研业务费专项资金项目 (2572019BF02)

收起

Detection method of hazelnut mildew based on hyperspectral feature fusion

  • ZHANG Dongyan ,
  • MAO Siyu ,
  • YANG Zijian ,
  • CHEN Nuo ,
  • WU Chenxu ,
  • MA Miaoyuan
Expand
  • (College of Computer and Control Engineering, Northeast Forestry University, Harbin 150040, China)

Received date: 2024-04-11

  Revised date: 2024-06-11

  Online published: 2025-02-14

Fold

摘要

为实现对榛子霉变的快速无损检测,研究将光谱特征与纹理特征融合并结合机器学习算法建立榛子霉变检测模型。采集400~1 000 nm的榛子样本高光谱图像,对样本的原始光谱使用标准正态变量变换法进行预处理,采用蜣螂优化算法、粒子群优化算法和连续投影算法3种特征波长选择方法对光谱进行特征选择;利用主成分分析法对高光谱图像进行降维,根据图像的贡献大小选择样本的最优主成分图像,结合灰度共生矩阵法提取样本4个角度上的5个纹理特征参数。分别基于样本光谱特征、纹理特征、光谱特征与纹理特征融合三类数据结合K最近邻算法构建榛子霉变检测模型。实验结果表明,基于蜣螂优化算法选择的特征光谱与纹理特征融合并结合K最近邻算法建立的模型效果最好,训练集和测试集准确率分别为99.20%和98.34%,实现了榛子霉变的快速无损检测。

关键词： 高光谱成像; 榛子; 霉变; 无损检测; 特征融合; 蜣螂优化算法

本文引用格式

张冬妍 , 毛思雨 , 杨子健 , 陈诺 , 吴晨旭 , 马苗源 . 基于高光谱特征融合的榛子霉变检测方法研究[J]. 食品与发酵工业, 2025 , 51(2) : 311 -319 . DOI: 10.13995/j.cnki.11-1802/ts.039523

Abstract

To realize the rapid and non-destructive detection of hazelnut mildew, the study fused spectral features with texture features, combining machine learning algorithms to establish a hazelnut mildew detection model.The hyperspectral images of hazelnut samples in the range of 400-1 000 nm were collected.The original spectrums were preprocessed with the standard normal variable transformation method.Dung beetle optimizer algorithm, particle swarm optimization algorithm, and successive projections algorithm were adopted to select characteristic wavelengths.The hyperspectral images were reduced the dimensionality with principal component analysis method, and the optimal principal component images of the samples were selected according to the contribution size of the images, utilizing the gray-level co-occurrence matrix method to extract five texture feature parameters on the four angles of samples.The hazelnut mildew detection K-nearest neighbor model was built based on spectral features, texture features, spectral features combined with texture features.Experimental results indicated that the best model was the K-nearest neighbor model under the fusion of texture features and spectral features selected by dung beetle optimizer algorithm.The accuracy of the model training set and test set were 99.20% and 98.34% respectively, realizing rapid and non-destructive detection of hazelnut mildew.

Key words： hyperspectral; hazelnut; mildew; non-destructive detection; feature fusion; dung beetle optimization algorithm

参考文献

[1] SALVATORE M M, ANDOLFI A, NICOLETTI R. Mycotoxin contamination in hazelnut: Current status, analytical strategies, and future prospects[J]. Toxins, 2023, 15(2):99.
[2] PANNICO A, SCHOUTEN R E, BASILE B, et al. Non-destructive detection of flawed hazelnut kernels and lipid oxidation assessment using NIR spectroscopy[J]. Journal of Food Engineering, 2015, 160:42-48.
[3] WU Q F, XIE L J, XU H R. Determination of toxigenic fungi and aflatoxins in nuts and dried fruits using imaging and spectroscopic techniques[J]. Food Chemistry, 2018, 252:228-242.
[4] WU Q F, XU H R. Detection of aflatoxin B1 in pistachio kernels using visible/near-infrared hyperspectral imaging[J]. Transactions of the ASABE, 2019, 62(5):1065-1074.
[5] MISHRA G, PANDA B K, RAMIREZ W A, et al. Application of SWIR hyperspectral imaging coupled with chemometrics for rapid and non-destructive prediction of Aflatoxin B1 in single kernel almonds[J]. LWT, 2022, 155:112954.
[6] ZOU Z Y, CHEN J, WANG L, et al. Nondestructive detection of peanuts mildew based on hyperspectral image technology and machine learning algorithm[J]. Food Science and Technology, 2022, 42: e71322.
[7] 孙钰莹, 章银, 沈飞, 等. 基于高光谱成像技术小麦籽粒霉变鉴别方法研究[J]. 中国粮油学报, 2022, 37(9):40-46.
SUN Y Y, ZHANG Y, SHEN F, et al. Identification method of wheat grain mildew based on hyperspectral imaging technology[J]. Journal of the Chinese Cereals and Oils Association, 2022, 37(9):40-46.
[8] 李凯楠, 沈广辉, 叶文武, 等. 基于近红外高光谱成像技术识别紫斑和霉变大豆的方法[J]. 南京农业大学学报, 2022, 45(4):691-698.
LI K N, SHEN G H, YE W W, et al. Identification method of soybean seeds with purple spot and moldy based on near infrared hyperspectral imaging technology[J]. Journal of Nanjing Agricultural University, 2022, 45(4):691-698.
[9] XU J Y, XU D C, BAI X P, et al. Non-destructive detection of moldy walnuts based on hyperspectral imaging technology[J]. Molecules, 2022, 27(20):6776.
[10] 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, 37(2):509-516.
SHEN G H, CAO Y Y, LIU X, et al. Identification of Fusarium damaged kernels using near infrared hyperspectral imaging and characteristic bands selection[J]. Jiangsu Journal of Agricultural Sciences, 2021, 37(2):509-516.
[11] 刘欢, 王雅倩, 王晓明, 等. 基于近红外高光谱成像技术的小麦不完善粒检测方法研究[J]. 光谱学与光谱分析, 2019, 39(1):223-229.
LIU H, WANG Y Q, WANG X M, et al. Study on detection method of wheat unsound kernel based on near-infrared hyperspectral imaging technology[J]. Spectroscopy and Spectral Analysis, 2019, 39(1):223-229.
[12] 殷勇, 王光辉. 连续投影算法融合信息熵选择霉变玉米高光谱特征波长[J]. 核农学报, 2020, 34(2):356-362.
YIN Y, WANG G H. Hyperspectral characteristic wavelength selection method for moldy maize based on continuous projection algorithm fusion information entropy[J]. Journal of Nuclear Agricultural Sciences, 2020, 34(2):356-362.
[13] 杨东, 王舒卉, 吴建华, 等. 玉米籽粒霉变等级高光谱图像检测方法研究[J]. 中国粮油学报, 2022, 37(11):46-53.
YANG D, WANG S H, WU J H, et al. Study on hyperspectral image detection method of maize grain mildew grade[J]. Journal of the Chinese Cereals and Oils Association, 2022, 37(11):46-53.
[14] 李斌, 苏成涛, 殷海, 等. 高光谱成像技术结合机器学习的稻米霉变检测[J]. 光谱学与光谱分析, 2023, 43(8):2391-2396.
LI B, SU C T, YIN H, et al. Hyperspectral imaging technology combined with machine learning for detection of moldy rice[J]. Spectroscopy and Spectral Analysis, 2023, 43(8):2391-2396.
[15] XUE J K, SHEN B. Dung beetle optimizer: A new meta-heuristic algorithm for global optimization[J]. The Journal of Supercomputing, 2023, 79(7):7305-7336.
[16] 戴天虹, 孙春雪, 黄建平, 等. 基于黄金正弦混沌斑鬣狗优化算法的高光谱波段选择[J]. 激光与光电子学进展, 2022, 59(10):519-528.
DAI T H, SUN C X, HUANG J P, et al. Hyperspectral wave band selection based on golden sine and chaotic spotted hyena optimizer algorithm[J]. Laser & Optoelectronics Progress, 2022, 59(10):519-528.
[17] 王艺婷, 黄世奇, 刘代志, 等. 基于粒子群优化算法的波段选择方法[J]. 计算机仿真, 2015, 32(9):417-420; 449.
WANG Y T, HUANG S Q, LIU D Z, et al. Band selection method based on particle swarm optimization[J]. Computer Simulation, 2015, 32(9):417-420; 449.
[18] 朱宗斌, 赵朝方, 曾侃, 等. 二进制粒子群支持向量机算法在SAR图像海面溢油特征选择的应用[J]. 海洋湖沼通报, 2015, 37(3):177-184.
ZHU Z B, ZHAO C F, ZENG K, et al. Oil spill feature selection in marine SAR images based on BPSO-SVM algorithm[J]. Transactions of Oceanology and Limnology, 2015, 37(3):177-184.
[19] 张冬妍, 付聪聪, 李丹丹, 等. 基于近红外光谱的榛子蛋白质无损检测模型[J]. 激光与光电子学进展, 2023, 60(1):401-407.
ZHANG D Y, FU C C, LI D D, et al. Nondestructive detection model of hazelnut protein based on near infrared spectroscopy[J]. Laser & Optoelectronics Progress, 2023, 60(1):401-407.
[20] LI B, ZOU J P, YIN H L, et al. Detection of the storage time of light bruises in yellow peaches based on spectrum and texture features of hyperspectral image[J]. Journal of Chemometrics, 2023, 37(11): e3516.
[21] 张冬妍, 马苗源, 黄莹, 等. 基于高光谱图像特征融合的榛子水分含量测定[J]. 现代食品科技, 2024,40(5):274-281.
ZHANG D Y, MA M Y, HUANG Y, et al. Determination of moisture content of hazelnuts based on hyperspectral image feature fusion[J]. Modern Food Science and Technology, 2024,40(5):274-281.
[22] 刘爽, 谭鑫, 刘成玉, 等. 高光谱数据处理算法的小麦赤霉病籽粒识别[J]. 光谱学与光谱分析, 2019, 39(11):3540-3546.
LIU S, TAN X, LIU C Y, et al. Recognition of Fusarium head blight wheat grain based on hyperspectral data processing algorithm[J]. Spectroscopy and Spectral Analysis, 2019, 39(11):3540-3546.
[23] 张玉荣, 邓毅浩, 褚洪强, 等. 基于高光谱成像技术的大豆霉变等级识别方法研究[J]. 河南工业大学学报(自然科学版), 2023, 44(6):105-113; 131.
ZHANG Y R, DENG Y H, CHU H Q, et al. Research on soybean mold level identification based on hyperspectral imaging technology[J]. Journal of Henan University of Technology (Natural Science Edition), 2023, 44(6):105-113; 131.
[24] 姜洪喆, 杨雪松, 李兴鹏, 等. 油茶果自然霉变程度的可见/近红外与中短波近红外光谱检测[J]. 食品科学, 2023, 44(4): 272-277.
JIANG H Z, YANG X S, LI X P, et al. Detection of the degree of natural mildew of Camellia oleifera fruit using visible/near Infrared, mid- and short-wave near infrared spectroscopy[J]. Food Science, 2023, 44(4): 272-277.
[25] MOSCETTI R, HAFF R P, AERNOUTS B, et al. Feasibility of Vis/NIR spectroscopy for detection of flaws in hazelnut kernels[J]. Journal of Food Engineering, 2013, 118(1):1-7.
Options
文章导航

/

〈 〉

技术支持：北京玛格泰克科技发展有限公司