基于CIW-YOLOv8n的棉花叶病害检测与识别方法

doi:10.14088/j.cnki.issn0439-8114.2025.10.032

湖北农业科学 ›› 2025, Vol. 64 ›› Issue (10): 207-212.doi: 10.14088/j.cnki.issn0439-8114.2025.10.032

基于CIW-YOLOv8n的棉花叶病害检测与识别方法

李佳骏, 董辉, 余霖, 刘金涛, 李双, 杨毅

云南农业大学大数据学院,昆明 650201

收稿日期:2025-02-03 出版日期:2025-10-25 发布日期:2025-11-14
通讯作者: 杨毅(1966-),男,云南昆明人,教授,主要从事农业信息化研究,(电子信箱)yyang66@126.com。
作者简介:李佳骏(2000-),男,云南文山人,在读硕士研究生,主要从事计算机视觉研究,(电子信箱)1123284638@qq.com。
基金资助:
云南省重大科技专项（202002AE09001002); 云南省高层次人才支持计划项目（YNWR-JXMS-2020-57）

A CIW-YOLOv8n-based method for cotton leaf disease detection and recognition

LI Jia-jun, DONG Hui, YU Lin, LIU Jin-tao, LI Shuang, YANG Yi

College of Big Data, Yunnan Agricultural University, Kunming 650201, China

Received:2025-02-03 Published:2025-10-25 Online:2025-11-14

摘要/Abstract

摘要： 针对棉花叶病害检测中存在的精确率低、检测效率不高的问题,从模型结构与检测效率两方面对YOLOv8模型进行改进,构建了CIW-YOLOv8n模型。在主干网络中引入C2f-ConvNeXtv2模块,该模块能根据输入图像动态调整卷积核,提升模型对不同图像内容的适应性与特征提取的准确性。在颈部网络中引入ImplicitHead检测头以替代原有结构,并将损失函数更换为WIoU,从而进一步优化模型的检测性能。结果表明,CIW-YOLOv8n模型的精确率、召回率、平均精度均值（mAP）、参数量、计算量分别为92.6%、82.4%、91.7%、2.761 MB、7.4×10⁹,该模型在保持较低复杂度的同时,实现了更高的检测精度,有效平衡了模型的性能与效率。

关键词: CIW-YOLOv8n模型, YOLOv8模型, 棉花叶病害, 检测, 识别

Abstract: To address the issues of low accuracy and insufficient detection efficiency in cotton leaf disease detection, the YOLOv8 model was improved in terms of both model structure and detection efficiency, leading to the construction of the CIW-YOLOv8n model.The C2f-ConvNeXtv2 module was introduced into the backbone network, which dynamically adjusted the convolution kernels based on the input image, thereby enhancing the model’ adaptability to diverse image content and the accuracy of feature extraction. The ImplicitHead detection head was introduced into the neck network to replace the original structure, and the loss function was replaced with WIoU, further optimizing the detection performance of the model.The results showed that the precision, recall, mean average precision (mAP), number of parameters, and computational cost of the CIW-YOLOv8n model were 92.6%, 82.4%, 91.7%, 2.761 MB, and 7.4×10⁹, respectively. This model achieved higher detection accuracy while maintaining low complexity, effectively balancing model performance and efficiency.

Key words: CIW-YOLOv8n model, YOLOv8 model, cotton leaf disease, detection, recognition

中图分类号:

TP391.4

李佳骏, 董辉, 余霖, 刘金涛, 李双, 杨毅. 基于CIW-YOLOv8n的棉花叶病害检测与识别方法[J]. 湖北农业科学, 2025, 64(10): 207-212.

LI Jia-jun, DONG Hui, YU Lin, LIU Jin-tao, LI Shuang, YANG Yi. A CIW-YOLOv8n-based method for cotton leaf disease detection and recognition[J]. HUBEI AGRICULTURAL SCIENCES, 2025, 64(10): 207-212.

参考文献

[1] 王晋伟,赵丽红,师勇强,等.棉花病害全程防治技术研究初报[J].中国棉花,2020,47(5):20-22,46.
[2] 戴建国,赖军臣.基于图像规则与Android手机的棉花病虫害诊断系统[J].农业机械学报,2015,46(1):35-44.
[3] BOCHKOVSKIY A, WANG C Y, LIAO H Y M. YOLOV4: Optimal speed and accuracy of obiect detection[J]. arXiv.2004.10934, 2020.
[4] WU Y J, YANG Y, WANG X F, et al.Fig fruit recognition method based on YOLOv4 deep learning[A].2021 18th international conference on electrical engineering/electronics, computer, telecommunications and information technology (ECTI-CON)[C]. Chiang Mai, Thailand:IEEE, 2021.
[5] REDMON J,FARHADI A.YOLO9000:Better,faster,stronger[A]. 2017 IEEE conference on computer vision and pattern recognition (CVPR)[C].Honolulu, HI, USA: IEEE, 2017. 6517-6525.
[6] 武星,齐泽宇,王龙军,等.基于轻量化YOLOv3卷积神经网络的苹果检测方法[J].农业机械学报,2020,51(8):17-25.
[7] 薛月菊,黄宁,涂淑琴,等.未成熟芒果的改进YOLOv2识别方法[J].农业工程学报,2018,34(7):173-179.
[8] FAN Y B, MAO S J, LI M, et al.CM-YOLOv8: Lightweight YOLO for coal mine fully mechanized mining face[J]. Sensors, 2024, 24(6): 1866.
[9] 郭文娟,冯全.基于改进YOLO的棉花叶片病害检测[J].干旱地区农业研究,2024,42(6):195-205.
[10] LI S, TAO T, ZHANG Y, et al.YOLO v7-CS: A YOLO v7-Based model for lightweight bayberry target detection count[J]. Agronomy, 2023, 13(12): 2952.
[11] 张楠楠,张晓,白铁成,等.基于CBAM-YOLO v7的自然环境下棉叶病虫害识别方法[J].农业机械学报,2023,54(S1):239-244.
[12] UDAWANT P, SRINATH P.Cotton leaf disease detection using instance segmentation[J]. Journal of cases on information technology, 2022, 24(4): 1-10.
[13] TRIPATHY S.Detection of cotton leaf disease using image processing techniques[J]. Journal of physics: Conference series, 2021(1): 012009.
[14] PAUL JOSHUA K,ALEX S A, MAGESWARI M, et al.Enhanced conditional self-attention generative adversarial network for detecting cotton plant disease in IoT-enabled crop management[J]. Wireless networks, 2025, 31(1): 299-313.
[15] WOO S, DEBNATH S, HU R H, et al.ConvNeXtv2: Co-designing and scaling convNets with masked autoencoders[A].2023 IEEE/CVF conference on computer vision and pattern recognition (CVPR)[C]. Vancouver, BC, Canada: IEEE, 2023.16133-16142.
[16] LIU Z, MAO H Z, WU C-Y, et al.A ConvNet for the 2020s[A].IEEE/CVF conference on computer vision and pattern recognition (CVPR)[C]. New Orleans,LA, USA: IEEE,2022.11966-11976.
[17] CHERPANATH E D, FATHIMA NASREEN P R, PRADEEP K, et al. Food image recognition and calorie prediction using faster R-CNN and mask R-CNN[A].9th international conference on smart computing and communications (ICSCC)[C]. Kochi, Kerala, India: IEEE, 2023. 83-89.
[18] EL GHAOUI L, GU F D, TRAVACCA B, et al.Implicit deep learning[J]. SIAM journal on mathematics of data science, 2021, 3(3): 930-958.
[19] TONG Z J, CHEN Y H, XU Z W, et al. Wise-IoU: Bounding box regression loss with dynamic focusing mechanism[J]. arXiv:2301.10051,2023.
[20] CHEN J R, KAO S H, HE H, et al.Run, don’t walk: Chasing higher FLOPS for faster neural networks[A]. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)[C]. Vancouver, BC, Canada: IEEE, 2023.12021-12031.
[21] WADEKAR S N, CHAURASIA A. MobileViTv3: Mobile-friendly vision transformer with simple and effective fusion of local, global and input features[J].arXiv:2209.15159,2022.

基于CIW-YOLOv8n的棉花叶病害检测与识别方法

A CIW-YOLOv8n-based method for cotton leaf disease detection and recognition

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	王春清, 尚书旗, 张茜雅, 刘伟, 岳丹松. 改进YOLOv8的步进式烤房烟叶烘烤阶段识别方法[J]. 湖北农业科学, 2025, 64(9): 202-212.
[2]	章磊, 冷欣, 陈佳凯, 李宗轩. 基于改进RT-DETR模型的油菜田间杂草识别研究[J]. 湖北农业科学, 2025, 64(8): 1-9.
[3]	李鹏飞, 曾靖. 基于改进YOLOv8的轻量化水稻病虫害识别模型研究[J]. 湖北农业科学, 2025, 64(8): 10-16.
[4]	贾惠言, 付岩, 王全胜, 吴银良. 杨梅中5种杀菌剂多残留检测方法比对[J]. 湖北农业科学, 2025, 64(7): 164-170.
[5]	霍静琦, 崔婷婷, 薛志璐. 基于改进YOLOv8模型的黄花菜花蕾识别研究[J]. 湖北农业科学, 2025, 64(7): 186-191.
[6]	谢鲜丽, 卿丽婷, 梁建丽, 刘丹, 李耀华. 瑶药猪殃殃化学成分、UPLC指纹图谱与化学模式识别评价[J]. 湖北农业科学, 2025, 64(6): 128-135.
[7]	杨爽, 周中林. 基于改进YOLOv8s-Seg模型的番茄成熟度检测[J]. 湖北农业科学, 2025, 64(6): 178-184.
[8]	张露, 吴雪莲. 基于改进YOLOv8模型的玉米叶斑病快速识别方法[J]. 湖北农业科学, 2025, 64(5): 160-166.
[9]	刘玉耀, 彭琼尹. 基于改进Faster R-CNN模型的丁岙杨梅成熟度检测方法[J]. 湖北农业科学, 2025, 64(4): 7-13.
[10]	崔建华, 李爱学, 吴瑞刚. 基于微信小程序平台的生物活性小分子检测服务微系统[J]. 湖北农业科学, 2025, 64(4): 19-23.
[11]	王志, 李寒冰, 王伟, 张玉杰, 田蜜, 王伟伟. 黄淮麦区小麦品种（系）品质与抗性基因的KASP分型检测[J]. 湖北农业科学, 2025, 64(4): 51-57.
[12]	黄徽, 郭银利, 彭青枝, 鲁振坦, 王福燕, 柳迪, 张苗. 基于糖基功能化磁性纳米颗粒材料富集-荧光法快速检测包装饮用水中铜绿假单胞菌[J]. 湖北农业科学, 2025, 64(2): 171-178.
[13]	赵小平, 董众祥, 冯瑾霞, 闫学龙, 胡晶洁. 基于改进NanoDet-Plus网络的树木小目标病虫害识别[J]. 湖北农业科学, 2025, 64(2): 192-196.
[14]	周洪, 孟小艳, 方伟舟, 丁晓晨. YOLOv8n-LF模型在机收小麦含杂率与破碎率检测中的应用[J]. 湖北农业科学, 2025, 64(10): 201-206.
[15]	张友昌, 张教海, 牛林涛, 王孝刚. 棉花病虫害智能识别技术研究进展[J]. 湖北农业科学, 2025, 64(10): 213-218.