[1]罗新磊,范菁.基于改进YOLO v8n的自然场景下苹果叶片病害检测[J].江苏农业科学,2025,53(20):288-296.
 Luo Xinlei,et al.Detection of apple leaf disease in natural environments based on improved YOLO v8n[J].Jiangsu Agricultural Sciences,2025,53(20):288-296.
点击复制

基于改进YOLO v8n的自然场景下苹果叶片病害检测()

《江苏农业科学》[ISSN:1002-1302/CN:32-1214/S]

卷:
第53卷
期数:
2025年第20期
页码:
288-296
栏目:
病虫害智能检测
出版日期:
2025-10-20

文章信息/Info

Title:
Detection of apple leaf disease in natural environments based on improved YOLO v8n
作者:
罗新磊范菁
云南民族大学电气信息工程学院/云南省无人自主系统重点实验室/云南省高校信息与通信安全灾备重点实验室,云南昆明 650500
Author(s):
Luo Xinleiet al
关键词:
图像识别深度学习病害检测YOLO v8n苹果叶片病害注意力机制多尺度特征
Keywords:
-
分类号:
S126;TP391.41
DOI:
-
文献标志码:
A
摘要:
针对苹果叶片病害特征不显著、相似度高等特点,为解决早期病斑检测难、人工检测效率低且准确率不高以及现有深度学习检测模型存在参数量大、检测精度差等问题,提出一种基于改进YOLO v8n的苹果叶片病害检测模型。该模型在骨干网络及颈部网络中使用基于AKConv模块改进的AKDown模块替代原有下采样模块,通过双通道可变核卷积提取病害特征,提升模型对不同尺度、不同显著性病害特征的识别能力。在骨干网络的最后设计改进的CBAM模块,利用多尺度卷积注意力与空间注意力捕捉不同尺寸的病斑特征及位置信息。利用加权双向特征金字塔(BiFPN)替代原有的颈部结构,通过跨层跳跃连接及特征加权自适应,实现多尺度特征高效融合。最后,引入GhostConv模块对C2f模块进行优化,生成虚拟通道减少卷积的通道数,有效减少模型的参数量。改进后的模型与原模型相比,参数量与浮点计算量下降24.2%、14.8%;黑星病的识别精度实现显著提升,较原模型精度提高5.7百分点;对5类不同苹果叶片病害检测的平均精度均值达93.5%。综上,本研究提出模型具有更优的检测精度和更小的参数量、浮点计算量,可为苹果叶片病害检测提供参考。
Abstract:
-

参考文献/References:

[1]中国苹果产业协会,国家苹果产业技术体系. 2022年苹果产业发展概况:一[J]. 中国果菜,2024,44(3):1-9.
[2]韩弘炜,张漪埌,齐立萍. 基于卷积神经网络的农作物病虫害检测综述[J]. 智慧农业导刊,2023,3(6):6-9.
[3]白雪松,吴建平,景文超,等. 基于卷积神经网络的农作物病虫害检测研究[J]. 计算机技术与发展,2022,32(12):200-205.
[4]温艳兰,陈友鹏,王克强,等.基于机器视觉的病虫害检测综述[J]. 中国粮油学报,2022,37(10):271-279.
[5]岳有军,刘杰琼,王红君,等. 基于改进YOLO v3模型的苹果树叶片病斑检测[J]. 中国科技论文,2021,16(11):1202-1208.
[6]刘斌,贾润昌,朱先语,等. 面向移动端的苹果叶部病虫害轻量级识别模型[J]. 农业工程学报,2022,38(6):130-139.
[7]晁晓菲,池敬柯,张继伟,等. 基于PSA-YOLO网络的苹果叶片病斑检测[J]. 农业机械学报,2022,53(8):329-336.
[8]Yu H L,Cheng X H,Chen C C,et al.Apple leaf disease recognition method with improved residual network[J]. Multimedia Tools and Applications,2022,81(6):7759-7782.
[9]Thapa,Ranjita,Zhang,et al. Plant Pathology 2021-FGVC8[EB/OL]. (2021-05-27)[2024-10-21]. https://kaggle.com/competitions/plant-pathology-2021-fgvc8.
[10]Pathological image of apple leaf[EB/OL]. (2021-08-18)[2024-10-21]. https://aistudio.baidu.com/aistudio/datasetdetail/11591.
[11]Jocher G,Qiu J ,Chaurasia A. Ultralytics YOLO (Version 8.0.0)[CP/OL]. [2024-10-21]. https://github.com/ultralytics/ultralytics.
[12]Dai J F,Qi H Z,Xiong Y W,et al. Deformable convolutional networks[C]//2017 IEEE International Conference on Computer Vision. Venice:IEEE,2017:764-773.
[13]Zhang X,Song Y,Song T,et al. AKConv:convolutional kernel with arbitrary sampled shapes and arbitrary number of parameters[EB/OL]. (2023-11-20)[2024-10-21]. https://doi.org/10.48550/arXiv.2311.11587.
[14]Woo S,Park J,Lee J Y,et al. CBAM:convolutional block attention module[C]//Computer Vision-ECCV 2018.Cham:Springer International Publishing,2018:3-19.
[15]Guo M H,Lu C Z,Hou Q,et al. Segnext:rethinking convolutional attention design for semantic segmentation[J].Advances in Neural Information Processing Systems,2022,35:1140-1156.
[16]Li H C,Xiong P F,An J,et al. Pyramid attention network for semantic segmentation[EB/OL]. (2018-05-25)[2024-10-21]. https://doi.org/10.48550/arXiv.1805.10180.
[17]Lin T Y,Dollár P,Girshick R,et al. Feature pyramid networks for object detection[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition. Honolulu:IEEE,2017:936-944.
[18]Tan M X,Pang R M,Le Q V. EfficientDet:scalable and efficient object detection[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle:IEEE,2020:10781-10790.
[19]Han K,Wang Y H,Tian Q,et al. GhostNet:more features from cheap operations[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle:IEEE,2020:1577-1586.

相似文献/References:

[1]张飞云.基于提升小波和学习向量量化神经网络的小麦病害图像识别[J].江苏农业科学,2013,41(05):103.
 Zhang Feiyun.Wheat diseases image recognition based on lifting wavelet and learning vector quantization neural network[J].Jiangsu Agricultural Sciences,2013,41(20):103.
[2]周洪刚,康敏.基于机器视觉的成熟柑橘自动识别研究[J].江苏农业科学,2013,41(06):380.
 Zhou Honggang,et al.Research on automatic recognizing of mature oranges based on machine vision[J].Jiangsu Agricultural Sciences,2013,41(20):380.
[3]刘丽娟,刘仲鹏.基于改进BP算法的玉米叶部病害图像识别研究[J].江苏农业科学,2013,41(11):139.
 Liu Lijuan,et al.Image recognition of maize leaf diseases based on improved BP algorithm[J].Jiangsu Agricultural Sciences,2013,41(20):139.
[4]刘丽娟,刘仲鹏.北方旱育稀植水稻病害图像识别预处理研究[J].江苏农业科学,2014,42(01):92.
 Liu Lijuan,et al.Study on image preprocessing of maize leaf diseases of dry-cultivated and sparse-planting rice in northern China[J].Jiangsu Agricultural Sciences,2014,42(20):92.
[5]何玲,陈长喜,许晓华.基于物联网的生猪屠宰监管系统关键技术研究[J].江苏农业科学,2017,45(06):201.
 He Ling,et al.Study on key technology of pig slaughtering supervision system based on internet of things[J].Jiangsu Agricultural Sciences,2017,45(20):201.
[6]梁万杰,曹宏鑫.基于卷积神经网络的水稻虫害识别[J].江苏农业科学,2017,45(20):241.
 Liang Wanjie,et al.Identification of rice insect pests based on CNN model[J].Jiangsu Agricultural Sciences,2017,45(20):241.
[7]刁智华,魏玉泉,刁春迎,等.基于图像的小麦白粉病形状特征参数优化与提取[J].江苏农业科学,2017,45(21):229.
 Diao Zhihua,et al.Image-based shape parameter optimization and extraction of wheat powdery mildew[J].Jiangsu Agricultural Sciences,2017,45(20):229.
[8]赵建敏,李艳,李琦,等.基于卷积神经网络的马铃薯叶片病害识别系统[J].江苏农业科学,2018,46(24):251.
 Zhao Jianmin,et al.Potato leaf disease identification system based on convolutional neural network[J].Jiangsu Agricultural Sciences,2018,46(20):251.
[9]李懿超,沈润平,黄安奇.基于深度学习的湘赣鄂地区植被变化及其影响因子关系模型[J].江苏农业科学,2019,47(03):213.
 Li Yichao,et al.Study on relational model between vegetation change and its impact factors based on deep learning in Hunan, Jiangxi and Hubei areas[J].Jiangsu Agricultural Sciences,2019,47(20):213.
[10]何彦虎,武传宇,童俊华,等.基于专家系统的穴盘苗品种识别算法设计与试验[J].江苏农业科学,2019,47(04):176.
 He Yanhu,et al.Design and experiment of identification algorithm of plug seedling based on expert system[J].Jiangsu Agricultural Sciences,2019,47(20):176.
[11]康飞龙,李佳,刘涛,等.多类农作物病虫害的图像识别应用技术研究综述[J].江苏农业科学,2020,48(22):22.
 Kang Feilong,et al.Application technology of image recognition for various crop diseases and insect pests: a review[J].Jiangsu Agricultural Sciences,2020,48(20):22.
[12]胡政,张艳,尚静,等.高光谱图像在农作物病害检测识别中的研究进展[J].江苏农业科学,2022,50(8):49.
 Hu Zheng,et al.Research progress of hyperspectral images in detection and identification of crop diseases[J].Jiangsu Agricultural Sciences,2022,50(20):49.
[13]路阳,刘婉婷,林立媛,等.CNN与BiLSTM相结合的水稻病害识别新方法[J].江苏农业科学,2023,51(20):211.
 Lu Yang,et al.A new method for rice disease identification by combining CNN and BiLSTM[J].Jiangsu Agricultural Sciences,2023,51(20):211.

备注/Memo

备注/Memo:
收稿日期:2024-10-21
基金项目:教育部-新一代信息技术创新项目(编号:2023IT077);云南省吴中海专家工作站项目(编号:202305AF150045)。
作者简介:罗新磊(1999—),男,江苏扬州人,硕士研究生,主要从事计算机视觉研究。E-mail:1341668312@qq.com。
通信作者:范菁,博士,教授,主要从事机器学习、模式识别、物联网研究。E-mail:798020566@qq.com。
更新日期/Last Update: 2025-10-20