高光谱成像的多种类柑橘病虫药害叶片检测方法研究

doi:10.3964/j.issn.1000-0593(2022)08-2397-06

摘要
参考文献
相关文章 (15)

全文: PDF (3184 KB)
输出: BibTeX | EndNote (RIS)

摘要：影响柑橘生长的病虫药害种类繁多，目前的检测方法大多针对单一病症，开发基于高光谱成像和机器学习的多种类柑橘病虫药害叶片快速精准检测方法，对果园精准施药和柑橘产业健康发展具有重要意义。以果园自然发病的柑橘叶片为研究对象，包括柑橘正常叶（50片）、溃疡病叶（50片）、煤烟病叶（103片）、缺素病叶（60片）、红蜘蛛叶（56片）和除草剂危害叶（85片），采集350～1 050 nm波段内的高光谱数据。分别利用一阶求导（1^stDer）、多元散射校正（MSC）和中值滤波（MF）方法对原始（Origin）高光谱数据进行预处理，对预处理后的高光谱数据采用主成分分析（PCA）和竞争性自适应重加权（CARS）算法提取特征波长，CARS降维得到的特征波长分别为10个、5个、12个和10个，4组PCA提取的特征波长均为7个，两种方法所得特征波长范围都集中在700～760 nm波段内。对全波段（FS）使用极限梯度提升树（XGBoost）算法，特征波长使用支持向量机（SVM）建立柑橘病叶多分类模型。采用XGBoost建立的检测识别模型有Origin-FS-XGBoost，1^stDer-FS-XGBoost，MSC-FS-XGBoost和MF-FS-XGBoost，对6种病虫害叶片检测得到的整体分类准确率（OA）分别为94.32%，93.60%，95.98%和96.56%；SVM建立的检测识别模型为Origin-CARS-SVM，1^stDer-CARS-SVM，MSC-CARS-SVM，MF-CARS-SVM，Origin-PCA-SVM，1^stDer-PCA-SVM，MSC-PCA-SVM和MF-PCA-SVM，各模型OA依次为93.63%，90.26%，87.90%，91.95%，87.53%，90.82%，83.50%和90.98%。结果表明，以FS为输入的XGBoost模型识别率整体优于以特征波长为输入的SVM模型，MF-FS-XGBoost模型OA为96.56%，召回率（Recall）为95.91%，模型训练时间（Train-time）为63 s，综合性能最好；CARS-SVM建模效果优于PCA-SVM，在3种预处理方式下，CARS-SVM模型识别率均高于87%，PCA-SVM模型识别率均在83%以上。结果证实了，高光谱成像技术结合机器学习方法可实现多种类柑橘病虫药害分类识别，为柑橘病虫药害快速无损检测和防治提供科学依据。

关键词：高光谱成像；柑橘病叶；特征波长提取；XGBoost；支持向量机

Abstract：Numerous kinds of diseases and insect pests affect citrus’ growth, but most of the current detection methods are for a single condition. It is important for the accurate application of pesticides in orchards and the healthy development of the citrus industry that the development of a detection method based on hyperspectral imaging and machine learning to achieve rapid and accurate detection of multiple pests and diseases on citrus leaves. Naturally onset citrus leaves in orchards were used as research objects, including normal citrus leaves (50 pieces), ulcer disease leaves (50 pieces), soot disease leaves (103 pieces), nutrient deficiency disease leaves (60 pieces), and red spider leaves (56 pieces) and herbicide damage leaves (85 pieces), hyperspectral data in the 350～1 050 nm band were collected. First-order derivation (1^stDer), multivariate scattering correction (MSC) and median filtering (MF) were used to preprocess the original (Origin) hyperspectral data, principal component analysis (PCA) and competitive adaptive weighting (CARS) algorithms were used to extract characteristic wavelengths from the prepossessed hyperspectral data. Characteristic wavelengths obtained by CARS were 10, 5, 12 and 10 respectively, and the 4 sets of characteristic wavelengths obtained by PCA were all 7, ranging in the 700～760 nm band. The limit gradient boosting tree (XGBoost) was used for the full band (FS), and the support vector machine (SVM) was used for the characteristic wavelength to establish a multi-classification model of citrus diseased leaves. The classification models established by XGBoost are Origin-FS-XGBoost, 1^stDer-FS-XGBoost, MSC-FS-XGBoost and MF-FS-XGBoost, and the overall classification accuracy (OA) obtained from the detection of 6 kinds of diseases and insect pests leaves was 94.32%, 93.60%, 95.98% and 96.56% respectively; the classification models established by SVM are Origin-CARS-SVM, 1^stDer-CARS-SVM, MSC-CARS-SVM, MF-CARS-SVM, Origin-PCA-SVM, 1^stDer-PCA-SVM, MSC-PCA-SVM and MF-PCA-SVM, model OA was 93.63%, 90.26%, 87.90%, 91.95%, 87.53%, 90.82%, 83.50% and 90.98% respectively. The experimental results demonstrate that the recognition rate of the XGBoost model with FS as input was better than the SVM model with characteristic wavelength as input. The OA of the MF-FS-XGBoost model was 96.56%, the Recall was 95.91%, and the model training time was 63 s. The overall performance was the best; the CARS-SVM modeling effect was better than PCA-SVM. After pre-processing by all three methods, the recognition rate of the CARS-SVM model was above 87%, and the recognition rate of the PCA-SVM model was above 83%. The results show that hyperspectral imaging technology combined with machine learning methods can classify and identify multiple species of citrus pests and diseases, providing a theoretical basis for rapid and non-destructive detection and control of citrus pests and diseases.

Key words：Hyperspectral imaging; Diseased citrus leaves; Characteristic wavelength extraction; XGBoost; SVM

收稿日期: 2021-11-29 修订日期: 2022-02-25

中图分类号:

O433.4

基金资助: 国家重点研发计划项目（2018YFC0807903）资助

作者简介: 吴叶兰，女，1970年生，北京工商大学人工智能学院副教授 e-mail: wu_yelan@163.com

引用本文:

吴叶兰，管慧宁，廉小亲，于重重，廖禺，高超. 高光谱成像的多种类柑橘病虫药害叶片检测方法研究[J]. 光谱学与光谱分析, 2022, 42(08): 2397-2402.
WU Ye-lan, GUAN Hui-ning, LIAN Xiao-qin, YU Chong-chong, LIAO Yu, GAO Chao. Study on Detection Method of Leaves With Various Citrus Pests and Diseases by Hyperspectral Imaging. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2022, 42(08): 2397-2402.

链接本文:

https://www.gpxygpfx.com/CN/10.3964/j.issn.1000-0593(2022)08-2397-06 或 https://www.gpxygpfx.com/CN/Y2022/V42/I08/2397

[1] YUE Xue-jun, LING Kang-jie, WANG Lin-hui, et al(岳学军，凌康杰，王林惠，等). Transactions of the Chinese Society for Agricultural Machinery(农业机械学报), 2019, 50(3): 186.
[2] Gu Q, Sheng L, Zhang T H, et al. Computers and Electronics in Agriculture, 2019, 167: 105066.
[3] Zhou X, Sun J, Tian Y, et al. International Journal of Remote Sensing, 2020, 41(6): 2263.
[4] GUO Wei, ZHU Yao-hui, WANG Hui-fang, et al(郭伟，朱耀辉，王慧芳，等). Transactions of the Chinese Society for Agricultural Machinery(农业机械学报), 2019, 50(9): 162.
[5] De Morsier F, Borgeaud M, Gas V, et al. IEEE Transactions on Geoscience & Remote Sensing, 2016, 54(6): 3410.
[6] Wang J, Wu W, Wang T, et al. Spectroscopy Letters, 2018, 51(9): 485.
[7] LAN Yu-bin, ZHU Zi-hao, DENG Xiao-ling, et al(兰玉彬，朱梓豪，邓小玲，等). Transactions of the Chinese Society of Agricultural Engineering(农业工程学报), 2019, 35(3): 92.
[8] DENG Xiao-ling, ZENG Guo-liang, ZHU Zi-hao, et al(邓小玲，曾国亮，朱梓豪，等). Journal of South China Agricultural University(华南农业大学学报), 2020, 41(6): 100.
[9] Abdulridha J, Batuman O, Ampatzidis Y. Remote Sensing, 2019, 11(11): 1373.
[10] LIU Yan-de, XIAO Huai-chun, SUN Xu-dong, et al(刘燕德，肖怀春，孙旭东，等). Transactions of the Chinese Society for Agricultural Machinery(农业机械学报), 2016, 47(11): 231.
[11] SUN Xu-dong, LIU Yan-de, XIAO Huai-chun, et al(孙旭东，刘燕德，肖怀春，等). Spectroscopy and Spectral Analysis(光谱学与光谱分析), 2017, 37(2): 551.
[12] Li H D, Liang Y Z, Xu Q S, et al. Analytica Chimca Acta, 2009, 648: 77.
[13] Tian X, Fan S X, Huang W Q, et al. Postharvest Biology and Technology, 2020, 161: 111071.
[14] DU Pei-jun, XIA Jun-shi, XUE Chao-hui, et al(杜培军，夏俊士，薛朝辉，等). Journal of Remote Sensing(遥感学报), 2016, 20(2): 236.
[15] Chen Tianqi, Guestrin C. XGBoost: A Scalable Tree Bossting System. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2016, 785.