[1]李萍,刘裕,师晓丽,等.基于多尺度残差空间注意力轻量化U-Net的农业害虫检测方法[J].江苏农业科学,2023,51(3):187-196.
 Li Ping,et al.Agriculture pest detection method based on multi-scale residual spatial attention lightweight U-Net[J].Jiangsu Agricultural Sciences,2023,51(3):187-196.
点击复制

基于多尺度残差空间注意力轻量化U-Net的农业害虫检测方法()

《江苏农业科学》[ISSN:1002-1302/CN:32-1214/S]

卷:
第51卷
期数:
2023年第3期
页码:
187-196
栏目:
农业工程与信息技术
出版日期:
2023-02-05

文章信息/Info

Title:
Agriculture pest detection method based on multi-scale residual spatial attention lightweight U-Net
作者:
李萍1 刘裕2 师晓丽1 张善文2
1.郑州西亚斯学院,河南郑州 451150; 2.西京学院信息工程学院,陕西西安 710123
Author(s):
Li Pinget al
关键词:
作物害虫检测U-Net空间注意力机制多尺度残差空间注意力轻量化U-Net
Keywords:
-
分类号:
TP391. 41
DOI:
-
文献标志码:
A
摘要:
田间害虫的快速精准检测是作物害虫防治的前提。现有基于卷积神经网络的作物害虫检测方法常包含大量训练参数,难以应用于现实场景中。针对上述难点,提出1种基于多尺度残差空间注意力轻量化U-Net(Multi-scale residual spatial attention lightweight U-Net,简称MSRSALU-Net)的检测方法,并应用于田间害虫检测。MSRSALU-Net由编码模块与解码模块组成。在MSRSALU-Net编码模块中,多尺度残差卷积模块用于提取害虫多尺度信息以缓解害虫尺度变化对检测性能的影响;空间注意力机制模块用于提取特征的全局依赖以缓解复杂背景对检测性能的干扰。此外,使用残差连接路径模块连接MSRSALU-Net的编码模块与解码模块,以更好地传播特征信息。在构建的IP13数据库上进行试验,基于MSRSALU-Net的害虫检测方法的识别精度为95.11%。与基于UNet、注意力UNet、MultiResUNet的害虫检测方法相比,MSRSALU-Net检测精度分别提高11.85%、5.38%、2.41%。模型参数量与U-Net、注意力UNet、MultiResUNet相比,分别减少了25.81%、21.45%、18.39%。结果表明,提出的MSRSALU-Net能有效克服害虫尺度变化、背景复杂等因素干扰,实现害虫的快速精准识别。该方法可为田间作物害虫检测系统提供技术支撑。
Abstract:
-

参考文献/References:

[1]Huo M Y,Tan J. Overview:research progress on pest and disease identification[M]//Pattern Recognition and Artificial Intelligence.Cham:Springer International Publishing,2020:404-415.
[2]Ayan E,Erbay H,Varn F,et al. Crop pest classification with a genetic algorithm-based weighted ensemble of deep convolutional neural networks[J]. Computers and Electronics in Agriculture,2020,179:105809.
[3]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.
[4]汪京京,张武,刘连忠,等. 农作物病虫害图像识别技术的研究综述[J]. 计算机工程与科学,2014,36(7):1363-1370.
[5]康飞龙,李佳,刘涛,等. 多类农作物病虫害的图像识别应用技术研究综述[J]. 江苏农业科学,2020,48(22):22-27.
[6]胡永强,宋良图,张洁,等. 基于稀疏表示的多特征融合害虫图像识别[J]. 模式识别与人工智能,2014,27(11):985-992.
[7]谢成军,李瑞,董伟,等. 基于稀疏编码金字塔模型的农田害虫图像识别[J]. 农业工程学报,2016,32(17):144-151.
[8]傅隆生,宋珍珍,Zhang Xin,等. 深度学习方法在农业信息中的研究进展与应用现状[J]. 中国农业大学学报,2020,25(2):105-120.
[9]Thenmozhi K,Srinivasulu Reddy U. Crop pest classification based on deep convolutional neural network and transfer learning[J]. Computers and Electronics in Agriculture,2019,164:104906.
[10]张博,张苗辉,陈运忠. 基于空间金字塔池化和深度卷积神经网络的作物害虫识别[J]. 农业工程学报,2019,35(19):209-215.
[11]Liu L,Wang R J,Xie C J,et al. PestNet:an end-to-end deep learning approach for large-scale multi-class pest detection and classification[J]. IEEE Access,7:45301-45312.
[12]Wang Q J,Zhang S Y,Dong S F,et al. Pest24:a large-scale very small object data set of agricultural pests for multi-target detection[J]. Computers and Electronics in Agriculture,2020,175(1):105585.
[13]Rahman C R,Arko P S,Ali M E,et al. Identification and recognition of rice diseases and pests using convolutional neural networks[J]. Biosystems Engineering,2020,194:112-120.
[14]钱蓉,孔娟娟,朱静波,等. 基于VGG-16卷积神经网络的水稻害虫智能识别研究[J]. 安徽农业科学,2020,48(5):235-238.
[15]乐毅,王文宇,张凯,等. 基于多层注意力机制的农业病虫害远程监督关系抽取研究[J]. 安徽农业大学学报,2020,47(4):682-686.
[16]Sweta P. Pedestrian detection based on hand-crafted features and multi-layer feature fused-ResNet model[J]. International Journal on Artificial Intelligence Tools,2021,30(5):2150027.
[17]He K,Zhang X,Ren S,et al. Deep residual learning for image recognition[C]. CVPR,2016:770-778.
[18]朱张莉,饶元,吴渊,等. 注意力机制在深度学习中的研究进展[J]. 中文信息学报,2019,33(6):1-11.
[19]Zhu H Y,Xie C,Fei Y Q,et al. Attention mechanisms in CNN-based single image super-resolution:a brief review and a new perspective[J]. Electronics,2021,10:1187.
[20]Li A,Chen J X,Kang B,et al. Adaptive multi-attention convolutional neural network for fine-grained image recognition[C]//2019 IEEE Globecom Workshops (GC Wkshps).December 9-13,2019.Waikoloa,HI,USA.IEEE,2019:1-5.
[21]Luo H F,Chen C C,Fang L N,et al. High-resolution aerial images semantic segmentation using deep fully convolutional network with channel attention mechanism[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing,2019,12(9):3492-3507.
[22]Zhang S H,Fu H Z,Yan Y G,et al. Attention guided network for retinal image segmentation[M]//Lecture notes in computer science. Cham:Springer International Publishing,2019:797-805.
[23]Huang T D,Fu S,Feng H N,et al. Bearing fault diagnosis based on shallow multi-scale convolutional neural network with attention[J]. Energies,2019,12(20):3937.
[24]张善文,邵彧,齐国红,等. 基于多尺度注意力卷积网络的作物害虫检测[J]. 江苏农业学报,2021,37(3):579-588.
[25]Sanjar K,Bekhzod O,Kim J,et al. Improved U-net:fully convolutional network model for skin-lesion segmentation[J]. Applied Sciences,2020,10(10):3658.
[26]Xu Z,Wang S,Stanislawski L V,et al. An attention U-Net model for detection of fine-scale hydrologic streamlines[J]. Environmental Modelling and Software,2021,140(5):104992.
[27]Zhao P,Zhang J,Fang W,et al. SCAU-Net:Spatial-channel attention U-Net for gland segmentation[J]. Frontiers in Bioengineering and Biotechnology,2020,8:670.
[28]Ali Y,Janabi-Sharifi F,Beheshti S. Echocardiographic image segmentation using deep Res-Unetwork[J]. Biomedical Signal Processing and Control,2021,64(8):102248.
[29]Wu X,Zhan C,Lai Y K,et al. IP102:A large-scale benchmark dataset for insect pest recognition[C]. CVPR,2019:8787-8796.
[30]Khalifa N E,Loey M,Taha M H N.Insect pests recognition based on deep transfer learning models[J]. Journal of Theoretical and Applied Information Technology,2020,98(1):60-68.
[31]Farahani A,Mohseni H.Medical image segmentation using customized U-Net with adaptive activation functions[J]. Neural Computing and Applications,2021,33(11):6307-6323.
[32]Ibtehaz N,Rahman M S.MultiResUNet:Rethinking the U-Net architecture for multimodal biomedical image segmentation[J]. Neural Networks,2020,121:74-87.

相似文献/References:

[1]亢洁,刘佳,刘文波,等.基于多感受野与多尺度融合的草莓病害分割算法[J].江苏农业科学,2024,52(4):206.
 Kang Jie,et al.Study on strawberry disease segmentation algorithm based on multi receptive field and multi-scale fusion module[J].Jiangsu Agricultural Sciences,2024,52(3):206.

备注/Memo

备注/Memo:
收稿日期:2022-03-28
基金项目:国家自然科学基金(编号:62072378)。
作者简介:李萍(1979—),女,河南郑州人,硕士,副教授,从事模式识别及其在精准农业大数据中的应用研究。E-mail:siasping@163.com。
通信作者:张善文,教授,博士生导师,研究方向为模式识别及其在作物病虫害检测中的应用。 E-mail:wjdw716@163.com。
更新日期/Last Update: 2023-02-05