广西师范大学学报（自然科学版） ›› 2022, Vol. 40 ›› Issue (6): 59-68.doi: 10.16088/j.issn.1001-6600.2022021601

• 研究论文 • 上一篇    下一篇

基于轻量级CNN的作物病虫害识别及安卓端应用

牛学德1, 高丙朋1*, 任荣荣2, 徐明明1   

  1. 1.新疆大学电气工程学院,新疆乌鲁木齐830047;
    2.四川信息职业技术学院智能控制学院,四川广元628040
  • 收稿日期:2022-02-16 修回日期:2022-03-09 出版日期:2022-11-25 发布日期:2023-01-17
  • 通讯作者: 高丙朋(1979—),男,新疆乌鲁木齐人,新疆大学副教授。E-mail:gbp_xd@sina.com
  • 基金资助:
    新疆维吾尔自治区自然科学基金(2019D01C079)

Crop Pestsand Diseases Identification and Android Application Based on Lightweight CNN

NIU Xuede1, GAO Bingpeng1*, REN Rongrong2, XU Mingming1   

  1. 1. School of Electrical Engineering, Xinjiang University, Urumqi Xinjiang 830047, China;
    2. Intelligent Control Academy, Sichuan Vocational College of Information Technology, Guangyuan Sichuan 628040, China
  • Received:2022-02-16 Revised:2022-03-09 Online:2022-11-25 Published:2023-01-17

PDF (PC)

63

摘要： 针对传统病虫害图像识别方法流程繁琐、效果差和应用困难等问题,本文以番茄、玉米、马铃薯3类作物17种叶部病虫害图片为研究对象,通过改进MobileNetV3网络模型并部署到移动端,实现了对多种作物病虫害图像的有效分类。首先,对病虫害图像做随机裁剪、旋转等预处理操作,对不均衡样本进行数据扩充;然后,将MobileNetV3网络从ImageNet数据集上学习获得的先验知识通过迁移学习策略应用到病虫害数据集上,经过参数微调并采用RAdam优化器训练后得到改进的轻量级网络模型;最后,将该模型通过Android Studio开发软件移植到安卓手机端。实验结果表明,该模型具有精度高、占用内存小、识别速度快等优势,能够满足对农作物叶片病虫害检测的基本要求,对智慧农业的发展具有参考意义。

关键词: 病虫害, 图像识别, 迁移学习, 轻量级, 安卓手机端

Abstract: Aiming at the problems of cumbersome process, poor effect and difficult application of traditional pest and disease image recognition methods, this paper takes pictures of 17 leaf diseases and pests of three crops of tomato, corn and potato as the research subjects. By improving the MobileNetV3 network model and deploying it to the mobile terminal, the effective classification of images of various crop diseases and insect pests is achieved. First, random cropping, rotation and other preprocessing operations are performed on the images of pests and diseases, and data expansion is performed on the unbalanced samples. Then the prior knowledge learned by the MobileNetV3 network from the ImageNet dataset is applied to the pests and diseases dataset through the transfer learning strategy. After fine-tuning the parameters and using the RAdam optimizer, the improved lightweight network model is obtained. Finally, the model is ported to the Android mobile phone through the Android Studio development software. The experimental results show that the model has the advantages of high precision, small memory occupation, and fast recognition speed, which can meet the basic requirements for detection of crop leaf diseases and insect pests, and has important reference significance for the development of smart agriculture.

Key words: diseases and pests, classification of images, transfer learning, lightweight, Android mobile phone

中图分类号: 

  • S431.9
[1] 李城恩, 潘晓映, 王美涵, 等.基于区间型数据计量的我国粮食产量研究[J].广西师范大学学报(自然科学版), 2022, 40(1): 206-215. DOI: 10.16088/j.issn.1001-6600.2021060914.
[2] 袁培森, 曹益飞, 马千里, 等.基于Random Forest的水稻细菌性条斑病识别方法研究[J].农业机械学报, 2021, 52(1): 139-145, 208. DOI: 10.6041/j.issn.1000-1298.2021.01.015.
[3] SZEGEDY C, LIU W, JIA Y Q, et al. Going deeper with convolutions[C]// 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).Los Alamitos, CA: IEEE Computer Society, 2015: 1-9. DOI: 10.1109/CVPR.2015.7298594.
[4] FERENTIONS K P.Deep learning models for plant disease detection and diagnosis[J].Computers and Electronics in Agriculture, 2018, 145: 311-318. DOI: 10.1016/j.compag.2018.01.009.
[5] ZHANG X H, QIAO Y, MENG F F, et al. Identification of maize leaf diseases using improved deep convolutional neural networks[J]. IEEE Access, 2018, 6: 30370-30377. DOI: 10.1109/ACCESS.2018.2844405.
[6] 陈善雄, 伍胜, 于显平, 等.基于卷积神经网络结合图像处理技术的荞麦病害识别[J].农业工程学报, 2021, 37(3): 155-163. DOI: 10.11975/j.issn.1002-6819.2021.03.019.
[7] 鲍文霞, 孙庆, 胡根生, 等.基于多路卷积神经网络的大田小麦赤霉病图像识别[J].农业工程学报, 2020, 36(11): 174-181. DOI: 10.11975/j.issn.1002-6819.2020.11.020.
[8] 任守纲, 贾馥玮, 顾兴健, 等.反卷积引导的番茄叶部病害识别及病斑分割模型[J].农业工程学报, 2020, 36(12): 186-195. DOI: 10.11975/j.issn.1002-6819.2020.12.023.
[9] 刘文波, 叶涛, 李颀.基于改进SOLO v2的番茄叶部病害检测方法[J].农业机械学报, 2021, 52(8): 213-220. DOI: 10.6041/j.issn.1000-1298.2021.08.021.
[10] KARTHIK R, HARIHARAN M, ANAND S, et al. Attention embedded residual CNN for disease detection in tomato leaves[J]. Applied Soft Computing, 2020, 86: 105933. DOI: 10.1016/j.asoc.2019.105933.
[11] KRISHNASWAMY RANGARAJAN A, PURUSHOTHAMAN R. Disease classification in eggplant using pre-trained VGG16 and MSVM[J].Scientific Reports, 2020, 10: 2322. DOI: 10.1038/s41598-020-59108-x.
[12] VERMA S, CHUG A, SINGH A P. Exploringcapsule networks for disease classification in plants[J].Journal of Statistics and Management Systems, 2020, 23(2): 307-315. DOI: 10.1080/09720510.2020.1724628.
[13] XIE B B, LIU J Z, HE M, et al. Research progress on autonomous navigation technology of agricultural robot[C]// 2021 IEEE 11th Annual International Conference on CYBER Technology in Automation, Control, and Intelligent Systems (CYBER).Piscatway, NJ: IEEE, 2021: 891-898. DOI: 10.1109/CYBER53097.2021.9588152.
[14] 李睿,王莹,王恒.基于树莓派的室内植物病虫害识别系统设计[J].电子设计工程,2022,30(8):114-118.DOI: 10.14022/j.issn1674-6236.2022.08.024.
[15] 苏博妮,化希耀,范振岐.基于移动互联网的水稻病虫害信息管理系统研究与设计[J].电子设计工程,2019,27(2):1-5.DOI: 10.14022/j.cnki.dzsjgc.2019.02.001.
[16] 孟亮, 郭小燕, 杜佳举, 等.一种轻量级CNN农作物病害图像识别模型[J].江苏农业学报, 2021, 37(5): 1143-1150. DOI: 10.3969/j.issn.1000-4440.2021.05.008.
[17] HOWARD A, SANDLER M, CHEN B, et al.Searching for MobileNetV3[C]// 2019 IEEE/CVF International Conference on Computer Vision (ICCV).Los Alamitos, CA: IEEE Computer Society, 2019: 1314-1324. DOI: 10.1109/ICCV.2019.00140.
[18] HU J, SHEN L, ALBANIE S, et al. Squeeze-and-excitation networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020, 42(8): 2011-2023. DOI: 10.1109/TPAMI.2019.2913372.
[19] 黄林生, 罗耀武, 杨小冬, 等.基于注意力机制和多尺度残差网络的农作物病害识别[J].农业机械学报, 2021, 52(10): 264-271. DOI: 10.6041/j.issn.1000-1298.2021.10.027.
[20] SANDLER M, HOWARD A, ZHU M L, et al. MobileNetV2: inverted residuals and linear bottlenecks[C]// 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.Los Alamitos, CA: IEEE Computer Society, 2018: 4510-4520. DOI: 10.1109/CVPR.2018.00474.
[21] PAN S J, YANG Q.A survey on transfer learning[J]. IEEE Transactions on Knowledge and Data Engineering, 2010, 22(10): 1345-1359. DOI: 10.1109/TKDE.2009.191.
[22] 张雪松, 庄严, 闫飞, 等.基于迁移学习的类别级物体识别与检测研究与进展[J].自动化学报, 2019, 45(7): 1224-1243. DOI: 10.16383/j.aas.c180093.
[23] 樊湘鹏, 许燕, 周建平, 等.基于迁移学习和改进CNN的葡萄叶部病害检测系统[J].农业工程学报, 2021, 37(6): 151-159. DOI: 10.11975/j.issn.1002-6819.2021.06.019.
[24] 王东方, 汪军.基于迁移学习和残差网络的农作物病害分类[J].农业工程学报, 2021, 37(4): 199-207. DOI: 10.11975/j.issn.1002-6819.2021.4.024.
[25] 徐建鹏, 王杰, 徐祥, 等.基于RAdam卷积神经网络的水稻生育期图像识别[J].农业工程学报, 2021, 37(8): 143-150. DOI: 10.11975/j.issn.1002-6819.2021.08.016.
[26] LIU L Y, JIANG H M, HE P C, et al. On the variance of the adaptive learning rate and beyond[EB/OL]. (2020-04-17)[2022-02-16]. https://arxiv.org/abs/1908.03265v3. DOI: 10.48550/arXiv.1908.03265.
[27] 黄建平, 陈镜旭, 李克新, 等.基于神经结构搜索的多种植物叶片病害识别[J].农业工程学报, 2020, 36(16): 166-173. DOI: 10.11975/j.issn.1002-6819.2020.16.021.
[28] 邓文轩, 杨航, 靳婷.基于注意力机制的图像分类降维方法[J].广西师范大学学报(自然科学版), 2021, 39(2): 32-40. DOI: 10.16088/j.issn.1001-6600.2020090704.
[1] 田晟, 宋霖. 基于CNN和Bagging集成的交通标志识别[J]. 广西师范大学学报（自然科学版）, 2022, 40(4): 35-46.
[2] 白捷, 高海力, 王永众, 杨来邦, 项晓航, 楼雄伟. 基于多路特征融合的Faster R-CNN与迁移学习的学生课堂行为检测[J]. 广西师范大学学报（自然科学版）, 2020, 38(5): 1-11.
[3] 严浩, 许洪波, 沈英汉, 程学旗. 开放式中文事件检测研究[J]. 广西师范大学学报（自然科学版）, 2020, 38(2): 64-71.
[4] 陈锦, 罗晓曙. 基于小波变换与野草算法的细胞图像特征提取与识别[J]. 广西师范大学学报（自然科学版）, 2015, 33(2): 22-28.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
版权所有 © 广西师范大学学报（自然科学版）编辑部
地址：广西桂林市三里店育才路15号　邮编：541004
电话：0773-5857325　E-mail: gxsdzkb@mailbox.gxnu.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发