[1]李豫晋,沈陆明,何少芳,等.基于改进MobileNet v3的苹果叶片病害识别研究[J].江苏农业科学,2024,52(12):224-231.
 Li Yujin,et al.Identification of apple leaf diseases based on improved MobileNet v3[J].Jiangsu Agricultural Sciences,2024,52(12):224-231.
点击复制

基于改进MobileNet v3的苹果叶片病害识别研究()

《江苏农业科学》[ISSN:1002-1302/CN:32-1214/S]

卷:
第52卷
期数:
2024年第12期
页码:
224-231
栏目:
农业工程与信息技术
出版日期:
2024-06-20

文章信息/Info

Title:
Identification of apple leaf diseases based on improved MobileNet v3
作者:
李豫晋沈陆明何少芳余文强滕明洪
湖南农业大学,湖南长沙 410125
Author(s):
Li Yujinet al
关键词:
苹果叶片病害识别MobileNet v3全维动态卷积ConvNext深度学习
Keywords:
-
分类号:
S436.611.1;TP391.41
DOI:
-
文献标志码:
A
摘要:
为解决移动端和嵌入式设备中苹果叶片病害识别准确率不高、效率低下的问题,提出了一种新的基于MobileNet v3网络的分类模型,以实现更加高效和准确的苹果叶片病害识别。首先通过数据增广方法增强数据集,按照9 ∶1的比例划分训练集和验证集;然后在MobileNet v3网络核心倒残差结构的升维部分引入全维动态卷积,以加强对不同维度注意力权重的学习,从而增强网络的拟合能力;最后在降维部分引入修改后的ConvNext Block模块,减少信息损失并增加全局感受野。采用PyTorch作为分类网络的深度学习框架,使用交叉熵损失函数作为分类任务的损失函数,Adam作为优化器,通过多组对比试验可知,MobileNet v1、MobileNet v2、ResNet34、MobileNet v3以及改进后的MobileNet v3 ODConvNext网络的准确率分别为94.5%、95.7%、97.2%、96.9%及97.5%。可见,MobileNet v3 ODConvNet网络拥有最高的Top-1准确率,相较于MobileNet v3网络和结构更为复杂的ResNet34网络分别提升了06、03百分点;在运算频率方面,相对于MobileNet v3网络仅增加了1.00×106次/s,并且仅为ResNet34网络参数量的1184%。因此,该试验结果证明了改进后的MobileNet v3 ODConvNext模型具有更加轻量级和更高准确率的优点,满足在移动端真实场景下进行苹果叶片病害识别的要求,有助于苹果叶片病害的防治工作。
Abstract:
-

参考文献/References:

[1] Gavhale M K R,Gawande P U. An overview of the research on plant leaves disease detection using image processing techniques[J]. IOSR Journal of Computer Engineering,2014,16(1):10-16.
[2]Kulkarni A H,Ashwin P R K. Applying image processing technique to detect plant diseases[J]. International Journal of Modern Engineering Research,2012,2 (5):3661-3664.
[3]王树桐,王亚南,曹克强. 近年我国重要苹果病害发生概况及研究进展[J]. 植物保护,2018,44(5):13-25,50.
[4]李保华,王彩霞,董向丽. 我国苹果主要病害研究进展与病害防治中的问题[J]. 植物保护,2013,39(5):46-54.
[5]Dubey S R,Jalal A S. Apple disease classification using color,texture and shape features from images[J]. Signal,Image and Video Processing,2016,10(5):819-826.
[6]Hinton G E,Salakhutdinov R R. Reducing the dimensionality of data with neural networks[J]. Science,2006,313(5786):504-507.
[7]le Cun Y,Bottou L,Bengio Y,et al. Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE,1998,86(11):2278-2324.
[8]le Cun Y,Boser B,Denker J S,et al. Backpropagation applied to handwritten zip code recognition[J]. Neural Computation,1989,1(4):541-551.
[9]杜岳峰,傅生辉,毛恩荣,等. 农业机械智能化设计技术发展现状与展望[J]. 农业机械学报,2019,50(9):1-17.
[10]郑文钟. 国内外智能化农业机械装备发展现状[J]. 现代农机,2015(6):4-8.
[11]姜红花,杨祥海,丁睿柔,等. 基于改进ResNet18的苹果叶部病害多分类算法研究[J]. 农业机械学报,2023,54(4):295-303.
[12]周宏威,沈恒宇,袁新佩,等. 基于迁移学习的苹果树叶片病虫害识别方法研究[J]. 中国农机化学报,2021,42(11):151-158.
[13]鲍文霞,吴刚,胡根生,等. 基于改进卷积神经网络的苹果叶部病害识别[J]. 安徽大学学报(自然科学版),2021,45(1):53-59.
[14]张云龙,袁浩,张晴晴,等. 基于颜色特征和差直方图的苹果叶部病害识别方法[J]. 江苏农业科学,2017,45(14):171-174.
[15]刘小玲,崔艳荣. 基于改进轻量化网络MobileNeXt的苹果叶片病害识别方法[J]. 江苏农业科学,2023,51(10):185-193.
[16]Sandler M,Howard A G,Zhu M L,et al. Inverted residuals and linear bottlenecks:mobile networks for classification,detection and segmentation[J]. CVPR,2018(6):4510-4520.
[17]Sandler M,Howard A,Zhu M L,et al. MobileNet v2:inverted ]residuals and linear bottlenecks[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake,2018:4510-4520.
[18]Howard A,Sandler M,Chen B,et al. Searching for MobileNet v3[C]//2019 IEEE/CVF International Conference on Computer Vision. Seoul,2019:1314-1324.
[19]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,2016:770-778.
[20]张宸嘉,朱磊,俞璐. 卷积神经网络中的注意力机制综述[J]. 计算机工程与应用,2021,57(20):64-72.
[21]Yang B,Bender G,Le Q V,et al. Condconv:conditionally parameterized convolutions for efficient inference[J]. Advances in Neural Information Processing Systems,2019,32.
[21]Yang B,Bender G,Le Q V,et al. CondConv:conditionally parameterized convolutions for efficient inference[EB/OL]. (2019-04-10)[2023-08-01].http://arxiv.org/abs/1904.04971.
[22]Li C,Zhou A J,Yao A B. Omni-dimensional dynamic convolution[EB/OL]. (2022-09-16)[203-08-01].http://arxiv.org/abs/2209.07947.
[23]Vaswani A,Shazeer N,Parmar N,et al. Attention is all you need[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems. Long Beach,2017:6000-6010.
[24]Liu Z,Mao H Z,Wu C Y,et al. A ConvNet for the 2020s[EB/OL]. (2022-01-10)[2023-08-01].http://arxiv.org/abs/2201.03545.

相似文献/References:

[1]徐凯宏,米雅婷,谷志新.基于GA-BP神经网络的温室番茄病害诊断[J].江苏农业科学,2016,44(04):387.
 Xu Kaihong,et al.Diagnosis of tomato disease in greenhouse based on GA-BP network[J].Jiangsu Agricultural Sciences,2016,44(12):387.
[2]张会敏,张云龙,张善文,等.基于区分矩阵的属性约简算法的作物病害识别方法[J].江苏农业科学,2015,43(01):387.
 Zhang Huimin,et al.A crop disease recognition method based on attribute reduction of discernibility matrix[J].Jiangsu Agricultural Sciences,2015,43(12):387.
[3]张云龙,袁浩,张晴晴,等.基于颜色特征和差直方图的苹果叶部病害识别方法[J].江苏农业科学,2017,45(14):171.
 Zhang Yunlong,et al.Apple leaf disease recognition based on color characteristics and differential histogram[J].Jiangsu Agricultural Sciences,2017,45(12):171.
[4]刁智华,刁春迎,魏玉泉,等.机器人系统中小麦病害识别与施药算法研究[J].江苏农业科学,2017,45(17):192.
 Diao Zhihua,et al.Study on wheat disease identification and spraying algorithm in robot system[J].Jiangsu Agricultural Sciences,2017,45(12):192.
[5]裴浩杰,冯海宽,李长春,等.基于多元线性回归和随机森林的苹果叶绿素含量高光谱估测方法比较[J].江苏农业科学,2018,46(17):224.
 Pei Haojie,et al.Comparison of hyperspectral estimation methods for chlorophyll contents of apple based on multiple linear regression and random forest[J].Jiangsu Agricultural Sciences,2018,46(12):224.
[6]刁智华,袁万宾,刁春迎,等.病害特征在作物病害识别中的应用研究综述[J].江苏农业科学,2019,47(05):71.
 Diao Zhihua,et al.Application of disease characteristics in crop disease identification:a review[J].Jiangsu Agricultural Sciences,2019,47(12):71.
[7]林彬彬,邱新法,何永健,等.茶树病害智能诊断识别算法研究[J].江苏农业科学,2019,47(06):85.
 Lin Binbin,et al.Study on intelligent diagnosis and recognition algorithm for tea diseases[J].Jiangsu Agricultural Sciences,2019,47(12):85.
[8]何前,郭峰林,方皓正,等.基于改进LeNet-5模型的玉米病害识别[J].江苏农业科学,2022,50(20):35.
 He Qian,et al.Study on maize disease recognition based on improved LeNet-5 model[J].Jiangsu Agricultural Sciences,2022,50(12):35.
[9]孙长兰,林海峰.一种基于集成学习的苹果叶片病害检测方法[J].江苏农业科学,2022,50(20):41.
 Sun Changlan,et al.An apple tree leaf disease detection method based on ensemble learning[J].Jiangsu Agricultural Sciences,2022,50(12):41.
[10]王慧,李康顺,蔡铁,等.基于约束性多目标优化算法的柑橘黄龙病识别算法[J].江苏农业科学,2023,51(6):159.
 Wang Hui,et al.Citrus Huanglongbing recognition algorithm based on constrained multi-objective optimization algorithm[J].Jiangsu Agricultural Sciences,2023,51(12):159.
[11]潘赟,陈松楠,马俊.基于多尺度特征融合Transformer的苹果叶片病害识别模型[J].江苏农业科学,2025,53(15):212.
 Pan Yun,et al.Apple leaf disease recognition model based on multiscale feature fusion Transformer[J].Jiangsu Agricultural Sciences,2025,53(12):212.

备注/Memo

备注/Memo:
收稿日期:2023-08-02
基金项目:湖南省自然科学基金(编号:2023JJ30304)。
作者简介:李豫晋(1998—),男,山西太原人,硕士研究生,主要从事农业信息技术研究。E-mail:531904940@qq.com。
通信作者:沈陆明,博士,教授,主要从事分形几何及其应用研究。E-mail:lum_s@126.com。
更新日期/Last Update: 2024-06-20