|本期目录/Table of Contents|

[1]戴硕,白涛,李东亚,等.基于知识蒸馏及改进ShuffleNet v2的棉花病虫害识别方法[J].江苏农业科学,2024,52(15):222-232.
 Dai Shuo,et al.Recognition of cotton pests and diseases based on knowledge distillation and improved ShuffleNet v2[J].Jiangsu Agricultural Sciences,2024,52(15):222-232.
点击复制

基于知识蒸馏及改进ShuffleNet v2的棉花病虫害识别方法(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15期
页码:
222-232
栏目:
农业工程与信息技术
出版日期:
2024-08-05

文章信息/Info

Title:
Recognition of cotton pests and diseases based on knowledge distillation and improved ShuffleNet v2
作者:
戴硕1白涛123李东亚1王震鲁1陈珍1
1.新疆农业大学计算机与信息工程学院,新疆乌鲁木齐 830052; 2.智能农业教育部工程研究中心,新疆乌鲁木齐 830052;3.新疆农业信息化工程技术研究中心,新疆乌鲁木齐 830052
Author(s):
Dai Shuoet al
关键词:
棉花病虫害ESSKNet知识蒸馏图像分类SK-AttentionShuffleNet v2
Keywords:
-
分类号:
TP391.41
DOI:
-
文献标志码:
A
摘要:
为探索及时、准确识别危害棉花叶片病虫害的方法,做好防护和治理工作。针对自然环境下棉花叶片图像受复杂背景影响导致分类精度降低以及模型参数量大使其不便于移动端部署的问题,提出了一种基于知识蒸馏的棉花病虫害识别模型ESSKNet。首先构建了包含8种类别的棉花病虫害图像数据集,其次通过在ShuffleNet v2模型引入SK-Attention自适应调整卷积核的大小关注棉花叶片不同尺寸大小的病斑信息并降低棉花病虫害图像复杂背景对模型的影响,将卷积核大小从3×3调整为5×5使神经网络更好地捕捉图像中的上下文信息和长程依赖关系。然后选取EfficientNet v2模型作为教师模型,ESSKNet模型作为学生模型,使用MGD方法进行知识蒸馏。试验结果表明,改进后的模型对棉花病虫害的识别准确率达96.06%,并且该模型参数量仅有EfficientNet v2的6.6%。该模型能有效识别棉花病虫害且更便于部署在移动设备上,以实现对棉花病虫害图像实时、精确地识别。
Abstract:
-

参考文献/References:

[1]张树珍. 棉花主要病害与防治措施[J]. 现代农业科技,2017(7):116-119.
[2]张红涛,朱洋,谭联,等. 基于FA-SVM技术的烟草早期病害识别[J]. 河南农业科学,2020,49(8):156-161.
[3]牛冲,牛昱光,李寒,等. 基于图像灰度直方图特征的草莓病虫害识别[J]. 江苏农业科学,2017,45(4):169-172.
[4]马慧琴,黄文江,景元书,等. 基于AdaBoost模型和mRMR算法的小麦白粉病遥感监测[J]. 农业工程学报,2017,33(5):162-169.
[5]张莉,纪铭阳,胡宗玉,等. 基于随机森林和逻辑回归分类模型的烟叶精选品控指标筛选[J]. 江苏农业科学,2020,48(3):214-217.
[6]任杰,柳毅,金鑫,等. 基于数据挖掘的蔬菜图片病害自动分类[J]. 北京师范大学学报(自然科学版),2006,42(6):592-596.
[7]王燕妮,贺莉. 基于多分类SVM的石榴叶片病害检测方法[J]. 计算机测量与控制,2020,28(9):191-195.
[8]Sugathan A,Sruthi S,Shamsudeen F M. A comparative study to detect rice plant disease using convolutional neural network (CNN) and support vector machine (SVM)[J]. Journal of Food,Agriculture & Environment,2020(2):18.
[9]Krizhevsky A,Sutskever I,Hinton G E. ImageNet classification with deep convolutional neural networks[J]. Communications of the ACM,2017,60(6):84-90.
[10]He K M,Zhang X Y,Ren S Q,et al. Deep residual learning for image recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition.Las Vegas:IEEE,2016:770-778.
[11]Szegedy C,Liu W,Jia Y Q,et al. Going deeper with convolutions[C]//2015 IEEE Conference on Computer Vision and Pattern Recognition.Boston:IEEE,2015:1-9.
[12]Li X,Rai L. Apple leaf disease identification and classification using ResNet models[C]//2020 IEEE 3rd International Conference on Electronic Information and Communication Technology (ICEICT).Shenzhen:IEEE,2020:738-742.
[13]王哲豪,范丽丽,何前. 基于MobileNet v2和迁移学习的番茄病害识别[J]. 江苏农业科学,2023,51(9):215-221.
[14]Vypirailenko D,Kiseleva E,Shadrin D,et al. Deep learning techniques for enhancement of weeds growth classification[C]//2021 IEEE International Instrumentation and Measurement Technology Conference (I2MTC).Glasgow:IEEE,2021:1-6.
[15]Nawaz M,Nazir T,Khan M A,et al. Plant disease classification using VGG-19 based faster-RCNN[C]//Singh M,Tyagi V,Gupta P,et al. International Conference on Advances in Computing and Data Sciences.Cham:Springer,2023:277-289.
[16]Ding X H,Zhang X Y,Han J G,et al. Scaling up your kernels to 31×31:revisiting large kernel design in CNNs[C]//2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition.New Orleans:IEEE,2022:11953-11965.
[17]Zeaiean Firrozabadi P,Hasani Moghaddam H. Evaluate the impact of Majority filter kernel size on increasing the accuracy of remotely sensed images[J]. Geography and Human Relationships,2018,1(2):806-817.
[18]Yim J,Joo D,Bae J,et al. A gift from knowledge distillation:fast optimization,network minimization and transfer learning[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).Honolulu:IEEE,2017:7130-7138.
[19]Yu X,Fu L R,Dai B S,et al. Soybean leaf morphology classification based on FPN-SSD and knowledge distillation[J]. Journal of Northeast Agricultural University,2020,27(4):9-17.
[20]Kurup R V,Anupama M A,Vinayakumar R,et al. Capsule network for plant disease and plant species classification[C]//Smys S,Tavares J,Balas V,et al. International conference on computational vision and bio inspired computing.Cham:Springer,2020:413-421.
[21]Li Y H,Pan Y W,Yao T,et al. Learning click-based deep structure-preserving embeddings with visual attention[J]. ACM Transactions on Multimedia Computing,Communications,and Applications,2019,15(3):1-19.
[22]Peng J L,Wang Y,Jiang P,et al. RiceDRA-net:precise identification of rice leaf diseases with complex backgrounds using a res-attention mechanism[J]. Applied Sciences,2023,13(8):4928.
[23]Yang B H,Gao Z W,Gao Y,et al. Rapid detection and counting of wheat ears in the field using YOLO v4 with attention module[J]. Agronomy,2021,11(6):1202.
[24]鲍浩,张艳. 基于注意力机制与改进残差模块的豆叶病害识别[J]. 江苏农业科学,2023,51(16):187-194.
[25]Shorten C,Khoshgoftaar T M. A survey on image data augmentation for deep learning[J]. Journal of Big Data,2019,6(1):60.
[26]Zhang L F,Bao C L,Ma K S. Self-distillation:towards efficient and compact neural networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2022,44(8):4388-4403.
[27]Yang Z D,Li Z,Shao M Q,et al. Masked generative distillation[M]. Cham:Springer Nature Switzerland,2022:53-69.
[28]Chen P G,Liu S,Zhao H S,et al. Distilling knowledge via knowledge review[C]//2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition.Nashville:IEEE,2021:5006-5015.
[29]Ma N N,Zhang X Y,Zheng H T,et al. ShuffleNet v2:practical guidelines for efficient CNN architecture design[C]//Computer Vision - ECCV 2018:15th European Conference,Munich:ACM,2018:122-138.
[30]Li X,Wang W,Hu X,et al. Selective kernel networks[C]//2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. California:IEEE,2019:510-519.
[31]McHugh M L. Interrater reliability:the kappa statistic[J]. Biochemia Medica,2012,22(3):276-282.

相似文献/References:

备注/Memo

备注/Memo:
收稿日期:2023-08-07
基金项目:新疆维吾尔自治区高校基本科研业务费科研项目(编号:XJEDU2022J009);中央引导地方科技发展专项(编号:ZYYD2022B12)。
作者简介:戴硕(2000—),男,安徽亳州人,硕士研究生,主要从事计算机视觉方向研究。E-mail:2402874453@qq.com。
通信作者:白涛,硕士,副教授,主要从事农业大数据、数据挖掘研究。E-mail:bt@xjau.edu.cn。
更新日期/Last Update: 2024-08-05