广西师范大学学报（自然科学版） ›› 2022, Vol. 40 ›› Issue (4): 104-114.doi: 10.16088/j.issn.1001-6600.2021120101

• 研究论文 • 上一篇    下一篇

基于SK-EfficientNet的番茄叶片病害识别模型

帖军1,2*, 隆娟娟1,2, 郑禄1,2, 牛悦1,2, 宋衍霖1,2   

  1. 1. 中南民族大学计算机科学学院,湖北武汉 430074;
    2. 湖北省制造企业智能管理工程技术研究中心,湖北 武汉 430074
  • 发布日期:2022-08-05
  • 通讯作者: 帖军(1976—), 男, 河南社旗人, 中南民族大学教授, 博士。E-mail: tiejun@mail.scuec.edu.cn
  • 基金资助:
    湖北省技术创新专项重大项目(2019ABA101); 湖北省科技重大专项(2020AEA011); 武汉市科技计划应用基础前沿项目(2020020601012267)

Tomato Leaf Disease Recognition Model Based on SK-EfficientNet

TIE Jun1,2*, LONG Juanjuan1,2, ZHENG Lu1,2, NIU Yue1,2, SONG Yanlin1,2   

  1. 1. School of Computer Science, South-Central Minzu University, Wuhan Hubei 430074, China;
    2. Hubei Provincial Engineering Research Center for Intelligent Management of Manufacturing Enterprises, Wuhan Hubei 430074, China
  • Published:2022-08-05

PDF (PC)

191

摘要： 针对目前番茄叶片病害识别的深度学习模型网络参数量多、精确度低、移动端模型部署难的问题,提出一种基于SK-EfficientNet的番茄叶片病害识别方法。该方法采用轻量级模型EfficientNet作为基准模型,并利用选择性卷积核机制SKNet替换EfficientNet核心模块MBConv中的SENet,使得卷积核根据输入特征的多尺度信息自适应选择感受野大小,提高图像特征提取能力同时更有效地利用参数。多组对比实验结果显示,改进后的模型在训练精度上得到进一步提高,且模型参数仅为3.83 MiB。在PlantVillage数据集上平均准确率达到99.64%,且验证SK-EfficientNet-B2的识别精度最高;在自然场景下平均准确率较原模型提高3.81个百分点。结果表明,改进后模型能有效提高自然场景下番茄叶片病害识别精度,可为移动端部署番茄叶片病害识别模型提供参考。

关键词: 番茄叶片, 病害识别, EfficientNet网络, SKNet, MBConv

Abstract: To solve the problems of the large amount of network parameters, low accuracy, and difficulty in deploying the mobile terminal model of the current deep learning model for tomato leaf diseased identification, a tomato leaf disease identification method based on SK-EfficientNet is proposed. The lightweight model EfficientNet is used as the benchmark model, and the selective convolution kernel mechanism SKNet is used to replace the SENet in the EfficientNet core module MBConv, so that the convolution kernel can adaptively chooses the size of the receptive field according to the multi-scale information of the input features. This method improves the ability to extract image features while making more effective use of parameters.Through multiple sets of comparative experiments, the results show that the training accuracy of the improved model has been further improved, and the model parameter is 3.83 MiB.The average accuracy rate on the PlantVillage data set is 99.64%, and the recognition accuracy of SK-EfficientNet-B2 is the highest; in natural scenes, the average accuracy rate is 3.81 percentage point higher than the original model.The results show that the improved model can effectively increase the accuracy of tomato leaf disease recognition in natural scenes, which provide a reference for the deployment of tomato leaf disease recognition models on mobile terminals.

Key words: tomato leaves, disease identification, EfficientNet network, SKNet, MBConv

中图分类号: 

  • TP391.41
[1] 柴毅, 黄席樾, 何离庆, 等. 番茄栽培病虫害防治知识表示[J]. 重庆大学学报(自然科学版), 2000, 23(6): 56-58. DOI: 10.11835/j.issn.1000-582X.2000.06.018.
[2]贾少鹏, 高红菊, 杭潇. 基于深度学习的农作物病虫害图像识别技术研究进展[J]. 农业机械学报, 2019, 50(增刊): 313-317. DOI: 10.6041/j.issn.1000-1298.2019.S0.048.
[3]王翔宇, 温皓杰, 李鑫星, 等. 农业主要病害检测与预警技术研究进展分析[J]. 农业机械学报, 2016, 47(9): 266-277. DOI: 10.6041/j.issn.1000-1298.2016.09.037.
[4]魏丽冉, 岳峻, 李振波, 等. 基于核函数支持向量机的植物叶部病害多分类检测方法[J]. 农业机械学报, 2017, 48(增刊): 166-171. DOI: 10.6041/j.issn.1000-1298.2017.S0.027.
[5]XIE C Q, HE Y. Spectrum and image texture features analysis for early blight disease detection on eggplant leaves[J]. Sensors, 2016, 16(5): 676. DOI: 10.3390/s16050676.
[6]柴阿丽, 李宝聚, 石延霞, 等. 基于计算机视觉技术的番茄叶部病害识别[J]. 园艺学报, 2010, 37(9): 1423-1430. DOI: 10.16420/j.issn.0513-353x.2010.09.030.
[7]刘君, 王学伟. 融合CNN多卷积特征与HOG的番茄叶部病害检测算法[J]. 北方园艺, 2020(4): 147-152. DOI: 10.11937/bfyy.20193405.
[8]唐熔钗, 伍锡如. 基于改进YOLO-V3网络的百香果实时检测[J]. 广西师范大学学报(自然科学版), 2020, 38(6): 32-39. DOI: 10.16088/j.issn.1001-6600.2020.06.004.
[9]THANGARAJ R, ANANDAMURUGAN S, PANDIYAN P, et al. Artificial intelligence in tomato leaf disease detection: a comprehensive review and discussion[J]. Journal of Plant Diseases and Protection, 2022, 129(3): 469-488. DOI:10.1007/s41348-021-00500-8.
[10]JIA S J, JIA P Y, HU S P,et al. Automatic detection of tomato diseases and pests based on leaf images[C]// 2017 Chinese Automation Congress (CAC). Piscataway NJ: IEEE Press,2017:3507-3510. DOI: 10.1109/CAC.2017.8243388.
[11]TM P, PRANATHI A, SAIASHRITHA K, et al. Tomato leaf disease detection using convolutional neural networks[C]// 2018 Eleventh International Conference on Contemporary Computing (IC3). Los Alamitos, CA: IEEE Computer Society, 2018: 314-318. DOI: 10.1109/IC3.2018.8530532.
[12]KAUR M, BHATIA R. Development of an improved tomato leaf disease detection and classification method[C]// 2019 IEEE Conference on Information and Communication Technology. Piscataway, NJ: IEEE Press, 2019: 1-5.DOI: 10.1109/ CICT48419.2019.9066230.
[13]GONZALEZ-HUITRON V, LEN-BORGES J A, RODRIGUEZ-MATA A E, et al. Disease detection in tomato leaves via CNN with lightweight architectures implemented in Raspberry Pi 4[J]. Computers and Electronics in Agriculture, 2021, 181: 105951. DOI: 10.1016/j.compag.2020.105951.
[14]郭小清, 范涛杰, 舒欣. 基于改进Multi-Scale AlexNet的番茄叶部病害图像识别[J]. 农业工程学报, 2019, 35(13): 162-169. DOI: 10.11975/j.issn.1002-6819.2019.13.018.
[15]汤文亮, 黄梓锋. 基于知识蒸馏的轻量级番茄叶部病害识别模型[J]. 江苏农业学报, 2021, 37(3): 570-578. DOI: 10.3969/j.issn.1000-4440.2021.03.004.
[16]张宁, 吴华瑞, 韩笑, 等. 基于多尺度和注意力机制的番茄病害识别方法[J]. 浙江农业学报, 2021, 33(7): 1329-1338. DOI: 10.3969/j.issn.1004-1524.2021.07.19.
[17]王美华, 吴振鑫, 周祖光. 基于注意力改进CBAM的农作物病虫害细粒度识别研究[J]. 农业机械学报, 2021, 52(4): 239-247. DOI: 10.6041/j.issn.1000-1298.2021.04.025.
[18]徐智,宁文昌,赵龙阳,等.基于注意力数据增广的细粒度图像分类方法[J]. 桂林电子科技大学学报,2021,41(6):496-503. DOI: 10.16725/j.cnki.cn45-1351/tn.2021.06.006.
[19]VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]// Advances in Neural Information Processing Systems 30 (NIPS 2017). Red Hook, NY: Curran Associates Inc., 2017: 6000-6010.
[20]LETARTE G, PARADIS F, GIGURE P, et al. Importance of self-attentionfor sentiment analysis[C]// Proceedings of the 2018 EMNLP Workshop BlackboxNLP: Analyzing and Interpreting Neural Networks for NLP. Stroudsburg, PA: Association for Computational Linguistics, 2018: 267-275. DOI: 10.18653/v1/W18-5429.
[21]YU C Q, WANG J B, PENG C, et al. BiSeNet: bilateral segmentation network for real-time semantic segmentation[C]// Computer Vision-ECCV 2018: Lecture Notes in Computer Science 11217. Cham: Springer Nature Switzerland AG, 2018: 334-349. DOI: 10.1007/978-3-030-01261-8_20.
[22]FU J L, ZHENG H L, MEI T. Look closer to see better: recurrent attention convolutional neural network for fine-grained image recognition[C]// 2017 IEEE Conference on Computer Vision and Pat-tern Recognition (CVPR).Los Alamitos, CA: IEEE Computer Society, 2017: 4476-4484. DOI: 10.1109/CVPR.2017.476.
[23]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.
[24]HE K M, ZHANG X Y, REN S Q, et al. Deep residual learning for image recognition[C]// 2016 the IEEE Conference on Computer Vision and Pattern Recognition(CVPR). Los Alamitos, CA: IEEE Computer Society, 2016: 770-778. DOI: 10.1109/CVPR.2016.90.
[25]TAN M X, LE Q V. EfficientNet: rethinking model scaling for convolutional neural networks[C]// Proceedings of the 36th International Conference on Machine Learning. Long Beach, CA: PMLR, 2019: 6105-6114.
[26]LI X, WANG W H, HU X L, et al. Selective kernel networks[C]// 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Los Alamitos, CA: IEEE Computer Society, 2019: 510-519. DOI: 10.1109/CVPR.2019.00060.
[27]唐浪, 李慧霞, 颜晨倩, 等. 深度神经网络结构搜索综述[J]. 中国图象图形学报, 2021, 26(2): 245-264. DOI: 10.11834/jig.200202.
[28]ZHANG T, QI G J, XIAO B, et al.Interleaved group convolutions[C]// 2017 IEEE International Conference on Computer Vision (ICCV). Los Alamitos, CA: IEEE Computer Society, 2017: 4383-4392. DOI: 10.1109/ICCV.2017.469.
[29]KRIZHEVSKY A, SUTSKEVER I, HINTON G E. ImageNet classification with deep convolutional neural networks[J].Communications of the ACM, 2017, 60(6): 84-90. DOI: 10.1145/3065386.
[30]CHEN L C, PAPANDREOU G, KOKKINOS l, et al. Deeplab: semantic image segmentation with deep convolutional nets, atrous convoution, and fully connected CRFs[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40(4): 834-848. DOI: 10.1109/TPAMI.2017.2699184.
[1] 万黎明, 张小乾, 刘知贵, 宋林, 周莹, 李理. 基于高效通道注意力的UNet肺结节CT图像分割[J]. 广西师范大学学报（自然科学版）, 2022, 40(3): 66-75.
[2] 张萍, 徐巧枝. 基于多感受野与分组混合注意力机制的肺结节分割研究[J]. 广西师范大学学报（自然科学版）, 2022, 40(3): 76-87.
[3] 李永杰, 周桂红, 刘博. 基于YOLOv3模型的人脸检测与头部姿态估计融合算法[J]. 广西师范大学学报（自然科学版）, 2022, 40(3): 95-103.
[4] 陈文康, 陆声链, 刘冰浩, 李帼, 刘晓宇, 陈明. 基于改进YOLOv4的果园柑橘检测方法研究[J]. 广西师范大学学报（自然科学版）, 2021, 39(5): 134-146.
[5] 张伟彬, 吴军, 易见兵. 基于RFB网络的特征融合管制物品检测算法研究[J]. 广西师范大学学报（自然科学版）, 2021, 39(4): 34-46.
[6] 张灿龙, 李燕茹, 李志欣, 王智文. 基于核相关滤波与特征融合的分块跟踪算法[J]. 广西师范大学学报（自然科学版）, 2020, 38(5): 12-23.
[7] 肖逸群, 宋树祥, 夏海英. 基于多特征的快速行人检测方法及实现[J]. 广西师范大学学报（自然科学版）, 2019, 37(4): 61-67.
[8] 王勋, 李廷会, 潘骁, 田宇. 基于改进模糊C均值聚类与Otsu的图像分割方法[J]. 广西师范大学学报（自然科学版）, 2019, 37(4): 68-73.
[9] 蔡冰, 张灿龙, 李志欣. 基于联合直方图的红外与可见光目标融合跟踪[J]. 广西师范大学学报（自然科学版）, 2017, 35(3): 37-44.
[10] 何鹏, 刘高凯, 李静辉. 基于机器视觉的疲劳驾驶监测预警系统[J]. 广西师范大学学报（自然科学版）, 2015, 33(4): 25-29.
[11] 王冬旭, 宋树祥, 蒋品群. 基于BP神经网络的竹片正反面识别算法[J]. 广西师范大学学报（自然科学版）, 2014, 32(2): 14-19.
[12] 马先兵, 孙水发, 覃音诗, 郭青, 夏平. 基于粒子滤波的on-line boosting目标跟踪算法[J]. 广西师范大学学报（自然科学版）, 2013, 31(3): 100-105.
[13] 孙水发, 李乐鹏, 董方敏, 邹耀斌, 陈鹏. 基于迭代阈值的子块部分重叠双直方图均衡算法[J]. 广西师范大学学报（自然科学版）, 2013, 31(3): 119-126.
[14] 唐振军, 戴玉敏, 张显全, 张师超. 基于DCT特征点的感知图像Hash函数[J]. 广西师范大学学报（自然科学版）, 2012, 30(3): 135-141.
[15] 薛其威, 伍锡如. 基于多模态特征融合的无人驾驶系统车辆检测[J]. 广西师范大学学报（自然科学版）, 2022, 40(2): 37-48.
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