基于改进YOLO11s的番茄叶片病害轻量化检测方法

Lightweight Detection Method for Tomato Leaf Diseases Based on Improved YOLO11s

摘要

针对现有番茄叶片病害检测中存在的背景复杂、模型参数量大以及小目标病害检测精度不足等问题，提出一种基于改进YOLO11s的番茄叶片病害轻量化检测方法。在主干网络中使用GSConv模块替换部分标准卷积模块，在保持检测精度的同时显著减少模型参数量；在特征提取层中引入 C3k2_PConv2模块，通过部分通道卷积运算有效减少冗余计算和内存访问量；在主干网络末端，将SENetV2模块与C2PSA模块融合，形成C2PSA_SENetV2模块，增强通道间信息交互和全局特征表达能力；引入UIoU损失函数，通过动态权重分配策略改善边界框回归精度。与原始YOLO11s模型相比，改进后的模型检测平均精度均值（mAP@0.5）提升了0.6% ，同时实现了显著轻量化，模型参数量减少了19.2% ，计算量降低了16.9% ，模型权重大小减少了19.1% 。本文提出的方法在保持检测精度的同时实现了显著的轻量化效果，可为番茄叶片病害实时智能检测提供技术支撑。

Abstract

A lightweight detection method for tomato leaf diseases based on improved YOLO11s is proposed to address the challenges of complex backgrounds，large model parameters，and insufficient detection accuracy for small-target diseases in existing tomato leaf disease detection systems. GSConv modules are employed to replace certain standard convolution modules in the backbone network.It significantly reduces the number of model parameters while maintaining detection accuracy.The C3k2_PConv2 module is introduced in the feature extraction layers to minimize redundant computations and memory access through partial channel convolution operations.At the end of the backbone network，the SENetV2 module is integrated with the C2PSA module to form the C2PSA_SENetV2 module，enhancing inter-channel information interaction and global feature representation capabilities.The UIoU loss function is incorporated to optimize bounding box regression accuracy through a dynamic weight allocation strategy.Compared to the original YOLO11s model，the improved model demonstrates a 0.6% improvement in mean average precision (mAP@0.5) while simultaneously achieving significant lightweight performance with decreases of 19.2% in model parameters，16.9% in computational complexity，and 19.1% in model weight size，respectively.The proposed method achieves substantial lightweight performance while maintaining detection accuracy，being able to provide technical support for real-time intelligent detection of tomato leaf diseases.

关键词

番茄叶片病害 / YOLO11s / 目标检测 / 轻量化

Key words

tomato leaf diseases / YOLO11s / object detection / lightweight

参考文献

[1] 贾京珠, 张天柱. 日光温室番茄典型病害及防治措施[J]. 现代园艺, 2020(1):173-174.

[2] Lecun Y, Bengio Y, Hinton G. Deep learning[J]. Nature, 2015, 521(7553):436-444.

[3] Girshick R, Donahue J, Darrell T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]//2014 IEEE Conference on Computer Vision and Pattern Recognition.Columbus,OH,USA:IEEE, 2014:580-587.

[4] Ren S Q, He K M, 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] Terven J, Córdova-Esparza D M, Romero-González J A. A comprehensive review of YOLO architectures in computer vision:from YOLOv1 to YOLOv8 and YOLO-NAS[J]. Machine Learning and Knowledge Extraction, 2023, 5(4):1680-1716.

[6] Liu W, Anguelov D, Erhan D, et al. SSD:single shot Multi- Box detector[C]//2016 European Conference on Computer Vision.Amsterdam,The Netherlands:Springer,2016:21-37.

[7] 赵子艺, 张德育, 徐双成. 基于改进YOLOv8的铁路异物检测算法[J]. 沈阳理工大学学报, 2025, 44(3):32-38,46.

[8] Zhao Z Y, Zhang D Y, Xu S C. Algorithm of railway foreign object intrusion detection based on improved YOLOv8[J]. Journal of Shenyang Ligong University, 2025, 44 (3):32-38,46. (in Chinese)

[9] 黄志龙. 基于改进YOLOv5s的番茄叶片病虫害识别方法[J]. 电大理工, 2024(3):1-7,17.

[10] 赵艳芹, 王萌. 改进YOLOv5模型的番茄叶部病害检测算法[J]. 黑龙江科技大学学报, 2023, 33(1):136-141.

[11] Zhao Y Q, Wang M. Detection of tomato leaf diseases based on improved YOLOv5 model[J]. Journal of Heilongjiang University of Science and Technology, 2023, 33 (1): 136-141. (in Chinese)

[12] 吴六爱, 许雪珂. 基于改进YOLOv10n的轻量化番茄叶片病虫害检测方法[J]. 智慧农业(中英文), 2025, 7(1):146-155.

[13] Wu L A, Xu X K. Lightweight tomato leaf disease and pest detection method based on improved YOLOv10n[J]. Smart Agriculture, 2025, 7(1):146-155. (in Chinese)

[14] 吴世琛, 毛玉明, 胡慧中, 等. 基于改进YOLO11n的葡萄果叶病害检测方法[J]. 农业工程学报, 2025, 41(14):140-147.

[15] Wu S C, Mao Y M, Hu H Z, et al. Detecting grape leaf diseases using improved YOLO11n[J]. Transactions of the Chinese Society of Agricultural Engineering, 2025, 41 (14): 140-147. (in Chinese)

[16] 黄铭杰, 蔡伟强, 张子健, 等. 基于改进YOLO11的荔枝果实品种实时精准识别算法[J]. 农业工程学报, 2025, 41 (11):156-164.

[17] Huang M J, Cai W Q, Zhang Z J, et al. Real-time accurate recognition algorithm for litchi fruit varieties based on improved YOLO11[J]. Transactions of the Chinese Society of Agricultural Engineering, 2025, 41(11):156-164. (in Chinese)

[18] Li H L, Li J, Wei H B, et al. Slim-neck by GSConv:a light-weight-design for real-time detector architectures[J]. Journal of Real-Time Image Processing, 2024, 21(3):62.

[19] Wang J, Zhao H. Improved YOLOv8 algorithm for water surface object detection[J]. Sensors, 2024, 24(15):5059.

[20] Zhang L Y, Wu X, Liu Z H, et al. ESD-YOLOv8 :an efficient solar cell fault detection model based on YOLOv8[J]. IEEE Access, 2024, 12:138801-138815.

[21] Liu Z G, Sun B S, Bi K Y. Optimization of YOLOv7 based on PConv,SE attention and wise-IoU[J]. International Journal of Computational Intelligence and Applications, 2024, 23: 2350033.

[22] Hughes D P, Salathé M. An open access repository of images on plant health to enable the development of mobile disease diagnostics[PP/OL]. arXiv(2016-04-12)[2025-08-15]. https://doi.org/10.48550/arXiv.1511.08060.

[23] Universitas Atma Jaya. Tomato leaf disease dataset[DS/OL]. Roboflow Universe(2023-04-01)[2025-08-15]. https://universe.roboflow.com/universitas-atma-jaya/tomato-leaf-disease-rxcft.

[24] Zhu X K, Lyu S C, Wang X, et al. TPH-YOLOv5:improved YOLOv5 based on transformer prediction head for object detection on drone-captured scenarios[C]//2021 IEEE/CVFInternational Conference on Computer Vision Workshops (ICCVW).Montreal,QC,Canada:IEEE,2021:2778-2788.

[25] Hussain M. YOLOv1 to v8:unveiling each variant:a comprehensive review of YOLO[J]. IEEE Access, 2024, 12:42816-42833.

[26] Murat A A, Kiran M S. A comprehensive review on YOLO versions for object detection[J]. Engineering Science and Technology,an International Journal, 2025, 70:102161.

[27] Ali M L, Zhang Z. The YOLO framework:a comprehensive review of evolution,applications,and benchmarks in object detection[J]. Computers, 2024, 13(12):336.

[28] Sapkota R, Flores-Calero M, Qureshi R, et al. YOLO advances to its genesis:a decadal and comprehensive review of the You Only Look Once(YOLO)series[J]. Artificial Intelligence Review, 2025, 58(9):274.

基金资助

国家引才引智示范基地项目(YZJD2023005)