|本期目录/Table of Contents|

[1]姜晟久,钟国韵.基于可分离扩张卷积和通道剪枝的番茄病害分类方法[J].江苏农业科学,2024,52(2):182-189.
 Jiang Shengjiu,et al.Tomato disease classification method based on separable dilated convolution and channel pruning[J].Jiangsu Agricultural Sciences,2024,52(2):182-189.
点击复制

基于可分离扩张卷积和通道剪枝的番茄病害分类方法(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2期
页码:
182-189
栏目:
农业工程与信息技术
出版日期:
2024-02-20

文章信息/Info

Title:
Tomato disease classification method based on separable dilated convolution and channel pruning
作者:
姜晟久钟国韵
东华理工大学信息工程学院,江西南昌 330013
Author(s):
Jiang Shengjiuet al
关键词:
番茄病害可分离扩张卷积通道剪枝MobileNet v2
Keywords:
-
分类号:
TP391.41
DOI:
-
文献标志码:
A
摘要:
为了实现番茄病害的快速检测,针对传统卷积神经网络病害分类方法参数量大、对算力要求高的问题,提出了一种基于可分离扩张卷积和通道剪枝的番茄病害分类方法。基于MobileNet v2,提出了一种可分离扩张卷积块,在不增加网络参数的情况下,扩大网络的感受野,提升网络提取番茄叶部病害特征的能力。然后替换PReLU激活函数,避免产生梯度弥散问题。同时能够更好地处理图像,提高网络对番茄叶部病害负值特征信息的提取能力,具有更好的鲁棒性。最后,使用通道剪枝技术,引入缩放因子联合权重损失函数,分辨相对不重要的通道,并对其进行裁剪,再对剪枝后的网络进行微调并重复以上步骤,在大幅减少网络参数量的同时,不影响网络的性能。在数据集上的结果表明,研究方法在网络参数量仅为0.7 M的情况下,准确率达到了96.44%,精确率达到了96.36%。与目前主流轻量化网络MobileNet v3、GhostNet、ShuffleNet v2相比,模型准确率分别提高了0.45、0.77、0.24百分点,同时模型参数量分别仅为以上模型的12.96%、13.46%、30.43%,模型更轻量且准确率更高。
Abstract:
-

参考文献/References:

[1]翟肇裕,曹益飞,徐焕良,等. 农作物病虫害识别关键技术研究综述[J]. 农业机械学报,2021,52(7):1-18.
[2]谭海文,吴永琼,秦莉,等. 我国番茄侵染性病害种类变迁及其发生概况[J]. 中国蔬菜,2019(1):80-84.
[3]Singh V K,Singh A K,Kumar A. Disease management of tomato through PGPB:current trends and future perspective[J]. 3 Biotech,2017,7(7):1-10.
[4]Luna R,Dadios E P,Bandala A A . Automated image capturing system for deep learning-based tomato plant leaf disease detection and recognition[C]//TENCON 2018-2018 IEEE Region 10 Conference. Jeju,South Korea. IEEE,2018:1414-1419.
[5]Zhang Y,Song C L,Zhang D W. Deep learning-based object detection improvement for tomato disease[J]. IEEE Access,2020,8:56607-56614.
[6]李超,彭进业,孔韦韦,等. 基于局部二值模式的作物叶部病斑检测[J]. 计算机工程与应用,2017,53(24):233-237.
[7]师韵,黄文准,张善文. 基于二维子空间的苹果病害识别方法[J]. 计算机工程与应用,2017,53(22):180-184.
[8]马浚诚,温皓杰,李鑫星,等. 基于图像处理的温室黄瓜霜霉病诊断系统[J]. 农业机械学报,2017,48(2):195-202.
[9]赵建敏,薛晓波,李琦. 基于机器视觉的马铃薯病害识别系统[J]. 江苏农业科学,2017,45(2):198-202.
[10]叶中华,赵明霞,贾璐. 复杂背景农作物病害图像识别研究[J]. 农业机械学报,2021,52(增刊1):118-124,147.
[11]王林柏,张博,姚竟发,等. 基于卷积神经网络马铃薯叶片病害识别和病斑检测[J]. 中国农机化学报,2021,42(11):122-129.
[12]顾兴健,朱剑峰,任守纲,等. 多尺度U网络实现番茄叶部病斑分割与识别[J]. 计算机科学,2021,48(增刊2):360-366,381.
[13]黄英来,艾昕. 改进残差网络在玉米叶片病害图像的分类研究[J]. 计算机工程与应用,2021,57(23):178-184.
[14]Saleem M H,Potgieter J,Arif K M. Plant disease detection and classification by deep learning[J]. Plants,2019,8(11):468.
[15]Dhiman P,Kukreja V,Manoharan P,et al. A novel deep learning model for detection of severity level of the disease incitrus fruits[J]. Electronics,2022,11(3):495.
[16]Guan X. A novel method of plant leaf disease detection based on deep learning and convolutional neural network[C]//2021 6th international conference on intelligent computing and signal processing. Xian,China:IEEE,2021:816-819.
[17]Loti N N A,Noor M R M,Chang S W. Integrated analysis of machine learning and deep learning in chili pest and disease identification[J]. Journal of the Science of Food and Agriculture,2020,101(9):3582-3594.
[18]Zhang Z. Improved Adam optimizer for deep neural networks[C]//2018 IEEE/ACM 26th international symposium on quality of service. Canada:ACM,2018:1-2.
[19]Sandler M,Howard A,Zhu M,et al. Mobilenet v2:inverted residuals and linear bottlenecks[C]//2018 IEEE/CVF conference on computer vision and pattern recognition. Salt Lake City,USA:IEEE,2018:4510-4520.
[20]Lei X Y,Pan H G,Huang X D. A dilated CNN model for image classification[J]. IEEE Access,2019,7:124087-124095.
[21]Wang P,Chen P,Yuan Y,et al. Understanding convolution for semantic segmentation[C]//2018 IEEE winter conference on applications of computer vision. Nevada,USA:IEEE,2018:1451-1460.
[22]姜晓勇,李忠义,黄朗月,等. 神经网络剪枝技术研究综述[J]. 应用科学学报,2022,40(5):838-849.
[23]Zhuang L,Li J G,Shen Z Q,et al. Learning efficient convolutional networks through network slimming[C]//Proceedings of the IEEE International Conference on Computer Vision. Venice,Italy,2017:2736-2744.
[24]He K M,Zhang X Y,Ren S Q,et al. Delving deep into rectifiers:surpassing human-level performance on imagenet classification[J]. IEEE Computer Society,2015:1026-1034.
[25]Howard A,Sandler M,Chen B,et al. Searching for Mobilenet v3[C]//2019 IEEE/CVF International conference on computer vision. South Korea:IEEE,2020:1314-1324.
[26]Ma N,Zhang X,Zheng H T,et al. Shufflenet v2:practical guidelines for efficient cnn architecture design[C]//Proceedings of the European conference on computer vision. Munich,Germany:IEEE,2018:116-131.
[27]Han K,Wang Y,Tian Q,et al. Ghostnet:more features from cheap operations[C]//2020 IEEE/CVF conference on computer vision and pattern recognition. Seattle,USA:IEEE,2020:1580-1589.

相似文献/References:

[1]严希,田山君,裴芸,等.几种中药提取液对番茄病害病原真菌的抑制效果[J].江苏农业科学,2017,45(20):129.
 Yan Xi,et al.Inhibitory effects of different Chinese herbs extracts against pathogenic fungi from tomato disease[J].Jiangsu Agricultural Sciences,2017,45(2):129.
[2]阎园园,陈华,姜波.基于群智能算法分类模型的番茄病害识别[J].江苏农业科学,2020,48(1):219.
 Yan Yuanyuan,et al.Study on classification algorithm of swarm intelligence algorithm in tomato disease identification[J].Jiangsu Agricultural Sciences,2020,48(2):219.
[3]倪智涛,胡伟健,李宝山,等.一种基于图像分类与目标检测协同的番茄细粒度病害识别方法[J].江苏农业科学,2023,51(22):221.
 Ni Zhitao,et al.A novel method for tomato fine-grained disease recognition based on image classification and target detection[J].Jiangsu Agricultural Sciences,2023,51(2):221.

备注/Memo

备注/Memo:
收稿日期:2023-03-06
基金项目:国家自然科学基金(编号:62162002);江西省主要学科学术和技术带头人领军人才项目(编号:20225BCJ22004)。
作者简介:姜晟久(2000—),男,湖南新化人,硕士研究生,研究方向为计算机视觉及其应用。E-mail:2021110251@ecut.edu.cn。
通信作者:钟国韵,博士,教授,硕士生导师,研究方向为计算机视觉、图像音视频处理。E-mail:gyzhong@ecut.edu.cn。
更新日期/Last Update: 2024-01-20