[1]施杰,林双双,张威,等.基于轻量化改进型YOLOv5s的玉米病虫害检测方法[J].江苏农业学报,2024,(03):427-437.[doi:doi:10.3969/j.issn.1000-4440.2024.03.005]
　SHI Jie,LIN Shuang-shuang,ZHANG Wei,et al.A corn disease and pest detection method based on lightweight improved YOLOv5s[J].,2024,(03):427-437.[doi:doi:10.3969/j.issn.1000-4440.2024.03.005]
点击复制

基于轻量化改进型YOLOv5s的玉米病虫害检测方法()
分享到：

江苏农业学报[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2024年03期
页码:
427-437
栏目:
植物保护
出版日期:
2024-03-30

文章信息/Info

Title:
A corn disease and pest detection method based on lightweight improved YOLOv5s
作者:
施杰1; 2; 林双双1; 张威1; 陈立畅1; 张毅杰1; 杨琳琳1; 2
(1.云南农业大学机电工程学院,云南昆明650201；2.云南省作物生产与智慧农业重点实验室,云南昆明650201)
Author(s):
SHI Jie1; 2; LIN Shuang-shuang1; ZHANG Wei1; CHEN Li-chang1; ZHANG Yi-jie1; YANG Lin-lin1; 2
(1.Faculty of Mechanical and Electrical Engineering, Yunnan Agricultural University, Kunming 650201, China；2.The Key Laboratory for Crop Production and Smart Agriculture of Yunnan Province, Kunming 650201, China)
关键词:
玉米; 病虫害; 检测模型; YOLOv5s; 轻量化
Keywords:
corn; pests and diseases; detection model; YOLOv5s; lightweight
分类号:
TP391.41;S432
DOI:
doi:10.3969/j.issn.1000-4440.2024.03.005
摘要:
针对复杂环境下目前现有的玉米病虫害检测方法的精度不理想、模型复杂、难以在移动端部署等问题,本研究提出了基于轻量化改进型YOLOv5s的玉米病虫害检测方法。首先,采用轻量级网络GhostNet替换原始YOLOv5s模型中特征提取网络和特征融合网络的卷积层,降低模型的计算量和参数量,提高运行速度,以满足移动端的部署要求；其次,为弥补GhostNet所带来的检测精度下降缺陷,在模型的主干特征提取网络中引入注意力机制,更加全面地评估特征权值,以增强玉米病虫害的特征,减弱无关信息的干扰,提升检测性能；最后,将模型的损失函数由CIOU替换为EIOU,以增强模型对目标的精确定位能力,从而提升模型的收敛速度和回归精度。试验结果表明,改进模型相比原始YOLOv5s模型在对供试玉米病虫害检测中，P、R和mAP分别提高了1.9个百分点、2.2个百分点和2.0个百分点,分别达到了94.6%、80.2%和88.8%；在保持较高检测精度的同时,模型的计算量、参数量和模型大小分别减少了50.6%、52.9%和50.4%,解决了检测模型在移动端的部署问题。
Abstract:
Aiming at the problems of unsatisfactory detection accuracy, complex model and difficult deployment on mobile terminals in the existing maize disease and pest detection methods in complex environments, this study proposed a maize disease and pest detection method based on lightweight improved YOLOv5s. Firstly, the lightweight network GhostNet was used to replace the convolutional layer in the feature extraction network and feature fusion network in the original YOLOv5s model, which reduced the calculation and parameter amount of the model and improved the running speed to meet the deployment requirements of the mobile terminal. Secondly, in order to compensate for the problem of detection accuracy degradation caused by GhostNet, the normalization-based attention module (NAM) was introduced into the backbone feature extraction network of the model to evaluate the feature weights more comprehensively, so as to enhance the characteristics of corn diseases and pests, weaken the interference of irrelevant information, and improve the detection performance. Finally, the loss function of the model was replaced by EIOU from CIOU to enhance the model’s ability to accurately locate the target, so as to improve the convergence speed and regression accuracy of the model. The experimental results showed that compared with the original YOLOv5s model, the P, R and mAP of the final improved model increased by 1.9 percentage points, 2.2 percentage points and 2.0 percentage points, respectively, reaching 94.6%, 80.2% and 88.8%. While maintaining high detection accuracy, the calculation amount, parameter amount and capacity of the model were reduced by 50.6%， 52.9% and 50.4%, which solved the deployment problem of the detection model on the mobile terminal.

参考文献/References:

[1]代瑞熙,徐伟平. 中国玉米增产潜力预测[J]. 农业展望,2022,18(3):41-49.
[2]王文明. 玉米常见病虫害防治措施探讨[J]. 农业开发与装备,2017(12):149-151.
[3]安艳丽. 玉米常见病害的识别与防治[J]. 农业与技术,2017,37(18):44.
[4]REN S, HE K, GIRSHICK R, et al. Faster R-CNN: towards real-time object detection with region proposal networks[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2016,39(6):1137-1149.
[5]THEODORIDIS S,KOUTROUMBAS K. Lecture notes in computer science[J]. Algorithmica,2016,540(8):169-195.
[6]陈桂珍,龚声蓉. 计算机视觉及模式识别技术在农业生产领域的应用[J]. 江苏农业科学,2015,43(8):409-413.
[7]黄丽明,王懿祥,徐琪,等. 采用YOLO算法和无人机影像的松材线虫病异常变色木识别[J]. 农业工程学报,2021,37(14):197-203.
[8]刘天真,滕桂法,苑迎春,等. 基于改进YOLOv3的自然场景下冬枣果实识别方法[J]. 农业机械学报,2021,52(5):17-25.
[9]FARJON G, KRIKEB O, HILLEL A B, et al. Detection and counting of flowers on apple trees for better chemical thinning decisions[J]. Precision Agriculture,2019,21(3):503-521.
[10]FERENTINOS K P. Deep learning models for plant disease detection and diagnosis[J]. Computers and Electronics in Agriculture,2018,145:311-318.
[11]苏俊楷,段先华,叶赵兵. 改进YOLOv5算法的玉米病害检测研究[J]. 计算机科学与探索,2023,17(4):933-941.
[12]徐会杰,黄仪龙,刘曼. 基于改进YOLOv3模型的玉米叶片病虫害检测与识别研究[J]. 南京农业大学学报,2022,45(6):1276-1285.
[13]周维,牛永真,王亚炜,等. 基于改进的YOLOv4-GhostNet水稻病虫害识别方法[J]. 江苏农业学报,2022,38(3):685-695.
[14]邢洁洁,谢定进,杨然兵,等. 基于YOLOv5s的农田垃圾轻量化检测方法[J]. 农业工程学报,2022,38(19):153-161.
[15]张振国,邢振宇,赵敏义,等. 改进YOLOv3的复杂环境下红花丝检测方法[J]. 农业工程学报,2023,39(3):162-170.
[16]王卫星,刘泽乾,高鹏,等. 基于改进YOLOv4的荔枝病虫害检测模型[J]. 农业机械学报,2023,54(5):227-235.
[17]陈玉冲,龙梦玲,郭辉,等. 灰飞虱研究的文献计量分析[J]. 江苏科技信息,2021,38(15):22-25,53.
[18]杨春媛. 玉米常见病害的发病症状及防治措施[J]. 乡村科技,2017(24):63-64.
[19]路兴涛,吴翠霞,张勇,等. 玉米灰飞虱与粗缩病的发生规律及综合防治策略[J]. 安徽农业科学,2013,41(19):8168-8169.
[20]REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once:unified,real-time object detection[C]. New York:IEEE,2016.
[21]REDMON J, FARHADI A. YOLOv3:an incremental improvement[J]. arXiv Preprint,2018.DOI:10.48550/arXiv.1804.02767.
[22]BOCHKOVSKIY A, WANG C Y, LIAO H Y M. YOLOv4:optimal speed and accuracy of object detection[J]. arXiv Preprint,2020.DOI:10.48550/arXiv.2004.10934.
[23]郭磊,王邱龙,薛伟,等. 基于改进YOLOv5的小目标检测算法[J]. 电子科技大学学报,2022,51(2):251-258.
[24]李文婧,徐国伟,孔维刚,等. 基于改进YOLOv4的植物叶茎交点目标检测研究[J]. 计算机工程与应用,2022,58(4):221-228.
[25]LIU Y C,SHAO Z R,TENG Y,et al. NAM:normalization-based attention module[J]. arXiv Preprint,2021.DOI:10.48550/arXiv.2111.12419.
[26]孔德鹏,常天庆,戴文君,等. 基于排序选择和精英引导的改进人工蜂群算法[J]. 控制与决策,2019,34(4):781-786.
[27]WOO S, PARK J, LEE J Y, et al. CBAM: convolutional block attention module[M]. Cham:Springer International Publishing,2018.
[28]YU J, JIANG Y, WANG Z, et al. Unitbox: an advanced object detection network[J]. ACM,2016.DOI:10.1145/2964284.2967274.
[29]ZHANG Y F, REN W, HANG Z, et al. Focal and efficient IOU loss for accurate bounding box regression[J]. Neurocomputing,2022,506:146-157.

相似文献/References:

[1]宝华宾,梁帅强,吕远大,等.玉米籽粒蛋白含量Meta-QTL及候选基因分析[J].江苏农业学报,2016,(04):736.[doi:10.3969/j.issn.100-4440.2016.04.004]
　BAO Hua-bin,LIANG Shuai-qiang,LYU Yuan- da,et al.Analysis of meta-QTL and candidate genes related to protein concentration in maize grain[J].,2016,(03):736.[doi:10.3969/j.issn.100-4440.2016.04.004]
[2]印志同,秦秋霞,阚欣,等.玉米快速叶绿素荧光参数全基因组关联分析[J].江苏农业学报,2016,(04):746.[doi:10.3969/j.issn.100-4440.2016.04.005]
　YIN Zhi-tong,QIN Qiu-xia,KAN Xin,et al.Genome-wide association analysis of fast chlorophyll fluorescence parameters in maize[J].,2016,(03):746.[doi:10.3969/j.issn.100-4440.2016.04.005]
[3]岳海旺,陈淑萍,彭海成,等.玉米籽粒灌浆特性品种间比较[J].江苏农业学报,2016,(05):1043.[doi:10.3969/j.issn.1000-4440.2016.05.014]
　YUE Hai-wang,CHEN Shu-ping,PENG Hai-cheng,et al.Grain filling characteristics in maize materials[J].,2016,(03):1043.[doi:10.3969/j.issn.1000-4440.2016.05.014]
[4]周玲,梁帅强,林峰,等.玉米二态性 InDel 位点的鉴定和分子标记开发[J].江苏农业学报,2016,(06):1223.[doi:doi:10.3969/j.issn.1000-4440.2016.06.005]
　ZHOU Ling,LIANG Shuai-qiang,LIN Feng,et al.Biallelic InDel loci detection and molecular marker development in maize[J].,2016,(03):1223.[doi:doi:10.3969/j.issn.1000-4440.2016.06.005]
[5]刘朝茂,李成云.玉米与大豆间作对玉米叶片衰老的影响[J].江苏农业学报,2017,(02):322.[doi:doi:10.3969/j.issn.1000-4440.2017.02.013]
　LIU Chao-mao,LI Cheng-yun.Effects of maize/soybean intercropping on maize leaf senescence[J].,2017,(03):322.[doi:doi:10.3969/j.issn.1000-4440.2017.02.013]
[6]江彬,毕银丽,申慧慧,等.氮营养与AM真菌协同对玉米生长及土壤肥力的影响[J].江苏农业学报,2017,(02):327.[doi:doi:10.3969/j.issn.1000-4440.2017.02.014]
　JIANG Bin,BI Yin-li,SHEN Hui-hui,et al.Synergetic effects of Arbuscular mycorrhizal fungus and nitrogen on maize growth and soil fertility[J].,2017,(03):327.[doi:doi:10.3969/j.issn.1000-4440.2017.02.014]
[7]李国锋,葛敏,吕远大.Opaque2转录因子对玉米α-醇溶蛋白基因家族成员表达的影响[J].江苏农业学报,2015,(06):1224.[doi:doi:10.3969/j.issn.1000-4440.2015.06.006]
　LI Guo-feng,GE Min,L Yuan-da.Differential expression of α-zein family genes regulated by Opaque2 transcription factor[J].,2015,(03):1224.[doi:doi:10.3969/j.issn.1000-4440.2015.06.006]
[8]刘飞,杨春艳,谢建新.傅里叶变换红外光谱结合判别分析法诊断蚕豆病虫害[J].江苏农业学报,2015,(03):531.[doi:10.3969/j.issn.1000-4440.2015.03.011]
　LIU Fei,YANG Chun-yan,XIE Jian-xin.Diagnosis of diseases and pests of broad bean by Fourier transform infrared spectroscopy combining discriminant analysis[J].,2015,(03):531.[doi:10.3969/j.issn.1000-4440.2015.03.011]
[9]管莉,张阿英.CaM 与 ZmCCaMK 相互作用参与 BR 诱导的玉米叶片抗氧化防护[J].江苏农业学报,2015,(01):10.[doi:10.3969/j.issn.1000-4440.2015.01.002]
　GUAN Li,ZHANG A-ying.CaM-ZmCCaMK interaction involved in brassinosteroid-induced antioxidant defense in leaves of maize[J].,2015,(03):10.[doi:10.3969/j.issn.1000-4440.2015.01.002]
[10]王元琮,何冰,林峰,等.调控玉米阻止授粉后叶片衰老的QTL定位[J].江苏农业学报,2017,(04):747.[doi:doi:10.3969/j.issn.1000-4440.2017.04.004]
　WANG Yuan-cong,HE Bing,LIN Feng,et al.QTL mapping for pollination-prevention on leaf senescence[J].,2017,(03):747.[doi:doi:10.3969/j.issn.1000-4440.2017.04.004]

备注/Memo

备注/Memo:
收稿日期：2023-05-18基金项目：国家自然科学基金项目(32260438)；云南省重大科技专项(202302AE09002002)；云南省作物生产与智慧农业重点实验室开放课题；云南省教育厅科学研究基金项目(2023Y0985)作者简介：施杰(1981-),男,云南昆明人,博士,副教授,主要从事智能农业装备、机械故障诊断技术研究。(E-mail)km_shijie@126.com通讯作者：杨琳琳,(E-mail)29545343@qq.com
更新日期/Last Update: 2024-05-20