|本期目录/Table of Contents|

[1]彭思绘,汪宇玲.基于P-MobileViT网络的小麦病害分类研究[J].江苏农业科学,2024,52(20):260-267.
 Peng Sihui,et al.Study on wheat disease classification based on P-MobileViT network[J].Jiangsu Agricultural Sciences,2024,52(20):260-267.
点击复制

基于P-MobileViT网络的小麦病害分类研究(PDF)
分享到:

《江苏农业科学》[ISSN:1002-1302/CN:32-1214/S]

卷:
第52卷
期数:
2024年第20期
页码:
260-267
栏目:
病虫害智能检测
出版日期:
2024-10-20

文章信息/Info

Title:
Study on wheat disease classification based on P-MobileViT network
作者:
彭思绘 汪宇玲
东华理工大学信息工程学院,江西南昌 330013
Author(s):
Peng Sihuiet al
关键词:
小麦病害病害分类MobileViT图像分割PoolFormer
Keywords:
-
分类号:
S126;TP391.41
DOI:
-
文献标志码:
A
摘要:
针对小麦病害图像分类方法的识别准确率不理想、模型参数量大等问题,提出一种基于P-MobileViT的小麦病害分类模型。首先对小麦图像进行健康和病害二分类,融合Grabcut算法、大律法对小麦病害图像的病斑区域进行分割;然后将病斑图像输入P-MobileViT分类模型,在其block的局部表征模块中引入深度卷积提取病斑图像的局部特征,在全局表征模块使用PoolFormer结构提取全局特征,以减少模型计算量和参数量;将输入特征图和全局特征叠加后与局部特征进行融合,从而强化模型对特征的分类能力。与经过迁移学习的轻量级深度学习模型MobileViT、ShuffleNet v2、MobileNet v3、GhostNet、EfficientNet v2在公开小麦病害数据集上进行试验对比,结果表明,P-MobileViT模型的准确率达到97.2%,比MobileViT模型高出了2.0百分点,同时参数量、推理时间分别减少了23.1%、31.6%;与其中准确率较高的模型MobileNet v3、GhostNet相比,P-MobileViT模型的准确率也分别提高3.1、3.3百分点,参数量分别减少58.3%、61.5%。在小麦病害分类任务中,P-MobileViT模型实现了识别精度的提升,有效减少了识别时间开销,且降低了模型复杂度。
Abstract:
-

参考文献/References:

[1]Liu W,Wang B,Zhao Z,et al. Historical review and countermeasures of wheat stripe rust epidemics in China[J]. China Plant Prot,2022,42:21-7.
[2]Picon A,Seitz M,Alvarez-Gila A,et al. Crop conditional Convolutional Neural Networks for massive multi-crop plant disease classification over cell phone acquired images taken on real field conditions[J]. Computers and Electronics in Agriculture,2019,167:105093.
[3]焦悦,王思曼,赵喜兰,等. 50个国外小麦品种(系)抗叶锈性鉴定[J]. 河南农业科学,2019,48(12):79-88.
[4]刁智华,袁万宾,罗雅雯,等. 基于图像处理的小麦白粉病病斑生长模型构建[J]. 中国农机化学报,2019,40(6):158-161.
[5]Boulent J,Foucher S,Théau J,et al. Convolutional neural networks for the automatic identification of plant diseases[J]. Frontiers in Plant Science,2019,10:941.
[6]孙俊,谭文军,毛罕平,等. 基于改进卷积神经网络的多种植物叶片病害识别[J]. 农业工程学报,2017,33(19):209-215.
[7]Xiong Y H,Liang L F,Wang L,et al. Identification of cash crop diseases using automatic image segmentation algorithm and deep learning with expanded dataset[J]. Computers and Electronics in Agriculture,2020,177:105712.
[8]Singh P,Verma A,Alex J S R. Disease and pest infection detection in coconut tree through deep learning techniques[J]. Computers and Electronics in Agriculture,2021,182:105986.
[9]顾博,邓蕾蕾,李巍,等. 基于GrabCut算法的玉米病害图像识别方法研究[J]. 中国农机化学报,2019,40(11):143-149.
[10]陈燕,朱成宇,胡小春,等. 基于改进Unet的小麦茎秆截面参数检测[J]. 农业机械学报,2021,52(7):169-176.
[11]冯晓,李丹丹,王文君,等. 基于轻量级卷积神经网络和迁移学习的小麦叶部病害图像识别[J]. 河南农业科学,2021,50(4):174-80.
[12]Nanni L,Manfè A,Maguolo G,et al. High performing ensemble of convolutional neural networks for insect pest image detection[J]. Ecological Informatics,2022,67:101515.
[13]Khan M K,Ullah M O. Deep transfer learning inspired automatic insect pest recognition[C]//Proceedings of the 3rd international conference on computational sciences and technologies. Jamshoro,Pakistan:Mehran University of Engineering and Technology,2022.
[14]Li C,Zhen T,Li Z H. Image classification of pests with residual neural network based on transfer learning[J]. Applied Sciences,2022,12(9):4356.
[15]杨非凡,徐伟诚,陈盛德,等. 融合Focal Loss与典型卷积神经网络结构的水稻病害图像分类[J]. 江苏农业科学,2023,51(14):198-204.
[16]惠巧娟,孙婕. 基于多尺度特征度量元学习的玉米叶片病害识别模型研究[J]. 江苏农业科学,2023,51(9):199-206.
[17]Khalifa N E M,Loey M,Taha M H N. Insect pests recognition based on deep transfer learning models[J]. J Theor Appl Inf Technol,2020,98(1):60-68.
[18]Pattnaik G,Shrivastava V K,Parvathi K. Transfer learning-based framework for classification of pest in tomato plants[J]. Applied Artificial Intelligence,2020,34(13):981-993.
[19]Mehta S,Rastegari M. MobileViT:light-weight,general-purpose,and mobile-friendly vision transformer[EB/OL]. (2021-10-05)[2023-11-01]. https://arxiv.org/abs/2110.02178.
[20]Vaswani A,Shazeer N,Parmar N,et al. Attention is all you need[EB/OL]. (2017-06-12)[2023-11-01]. https://arxiv.org/abs/1706.03762.
[21]Dosovitskiy A,Beyer L,Kolesnikov A,et al. An image is worth 16×16 words:transformers for image recognition at scale[EB/OL]. (2020-10-22)[2023-11-01]. https://arxiv.org.abs/2010.11929.
[22]Howard A,Sandler M,Chen B,et al. Searching for MobileNet v3[C]//2019 IEEE/CVF International Conference on Computer Vision (ICCV).Seoul,Korea (South):IEEE,2019:1314-1324.
[23]Touvron H,Cord M,Douze M,et al. Training data-efficient image transformers & distillation through attention[EB/OL]. (2020-12-23)[2023-11-01]. https://arxiv.org/abs/2012.12877.
[24]Howard A G,Zhu M L,Chen B,et al. MobileNets:efficient convolutional neural networks for mobile vision applications[EB/OL]. (2017-04-17)[2023-11-01]. https://arxiv.org/abs/1704.04861.
[26]Otsu N. A threshold selection method from gray-level histograms[J]. IEEE Transactions on Systems,Man,and Cybernetics,1979,9(1):62-66.
[27]Rother C,Kolmogorov V,Blake A. GrabCut:interactive foreground extraction using iterated graph cuts[M]//Seminal Graphics Papers:Pushing the Boundaries,Volume 2.New York,NY,USA:ACM,2023:593-598.

相似文献/References:

[1]张飞云.基于提升小波和学习向量量化神经网络的小麦病害图像识别[J].江苏农业科学,2013,41(05):103.
 Zhang Feiyun.Wheat diseases image recognition based on lifting wavelet and learning vector quantization neural network[J].Jiangsu Agricultural Sciences,2013,41(20):103.

备注/Memo

备注/Memo:
收稿日期:2023-11-24
基金项目:国家自然科学基金(编号:62066003);国家留学基金(编号:CSC202208360143);江西省研究生创新专项资金(编号:YC2022-s626)。
作者简介:彭思绘(1997—),女,安徽池州人,硕士研究生,主要研究方向为计算机视觉。E-mail:pengsh1002@163.com。
通信作者:汪宇玲,博士,教授,硕士生导师,主要从事模式识别与图像处理、计算机视觉研究。E-mail:wangyuling_119@vip.163.com。
更新日期/Last Update: 2024-10-20