Research and implementation of a lightweight YOLOv5 for apple leaf disease detection

doi:10.13733/j.jcam.issn.2095-5553.2025.04.019

Abstract

Abstract: With the continuous improvement of agricultural intelligence level, the automatic detection method of apple leaf disease becomes important and necessary, but the existing automatic detection algorithm model is often difficult to deploy on the mobile terminal due to the complex network structure, a lightweight model of YOLOv5—SCFG is proposed in this paper. Firstly, the lightweight ShuffleNetv2 network is introduced to rebuild the YOLOv5 backbone network, so as to ensure the overall lightweight network. Additionally, the CARAFE up‑sampling operator and FasterNet module are introduced into the neck network, so as to enhance feature extraction capabilities and speed up feature fusion. Finally, the global attention mechanism (GAM) is introduced to compensate for the loss of precision caused by network lightweighting. The experimental results show that the weight size of YOLOv5—SCFG model is 6.2 MB, mAP is 85.9%, and FLOPs is 6.3 G. Compared with YOLOv5s, the weight size of YOLOV5-SCFG model is reduced by 57%, mAP is decreased by 0.2%, and FLOPs is decreased by 61%.

Key words: apple leaves, lightweight model, disease detection, YOLOv5s

摘要： 随着农业智能化水平不断提升，苹果叶片病害自动化检测十分必要，而现有的自动化检测模型由于网络结构复杂，难以在移动端进行部署。基于此，构建一种YOLOv5—SCFG轻量模型。首先，引入轻量型网络ShuffleNetv2重新构建YOLOv5骨干网络,保证网络整体轻量化；然后，在颈部网络引入CARAFE上采样算子和FasterNet模块，增强特征提取能力，加快特征融合速度；最后，添加全局注意力机制GAM，弥补网络轻量化带来的精度损失。结果表明，YOLOv5—SCFG模型权重大小为6.2 MB，平均精度均值mAP为85.9%，计算量FLOPs为6.3 G。相比于YOLOv5s，模型权重大小减少57%，mAP下降0.2%，FLOPs减小61%。

关键词: 苹果叶片, 轻量化模型, 病害检测, YOLOv5s

CLC Number:

Zhou Junchang, Zeng Wei, , Peng Peng, , Pang Jicheng, Liu Junjun, Yang Xilin. Research and implementation of a lightweight YOLOv5 for apple leaf disease detection[J]. Journal of Chinese Agricultural Mechanization, 2025, 46(4): 126-132.

周俊昌, 曾维, 彭鹏, 庞记成, 刘军军, 杨熙临. 苹果叶片病害检测的轻量化YOLOv5研究与实现[J]. 中国农机化学报, 2025, 46(4): 126-132.

References

[ 1 ] Khan A I, Quadri s MK, Banday S, et al. Deep diagnosis: A real‑time apple leaf disease detection system based on deep learning [J]. Computers and Electronics in Agriculture, 2022, 198: 107093.
[ 2 ] Ding R, Qiao Y, Yang X, et al. Improved resnet based apple leaf diseases identification [J]. IFAC‑PapersOnLine, 2022, 55(32): 78-82.
[ 3 ] Dutot M, Nelson L M, Tyson R C. Predicting the spread of postharvest disease in stored fruit, with application to apples [J]. Postharvest Biology and Technology, 2013, 85: 45-56.
[ 4 ] Sarker I H. Deep learning: A comprehensive overview on techniques, taxonomy, applications and research directions [J]. SN Computer Science, 2021, 2(6): 420.
[ 5 ] Girshick R, Donahue J, Darrell T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation [C]. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2014: 580-587.
[ 6 ] Girshick R. Fast R—CNN [C]. 2015 IEEE International Conference on Computer Vision (ICCV) 2015: 1440-1448.
[ 7 ] Ren S, He K, Girshick R, et al. Faster R—CNN: Towards real‑time object detection with region proposal networks [J]. IEEE Transactions Pattern Analysis and Machine Intelligence, 2017, 39(6): 1137-1149.
[ 8 ] Redmon J, Divvala S, Girshick R, et al. You only look once: Unified, real‑time object detection [C]. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016: 779-788.
[ 9 ] Liu W, Anguelov D, Erhan D, et al. SSD: Single shot multibox detector [C]. Computer Vision‑ECCV 2016: 14th European Conference, Amsterdam, the Netherlands, October 11-14, 2016, Proceedings, part I 14. Springer International Publishing, 2016: 21-37.
[10] Lin T Y, Goyal P, Girshick R, et al. Focal loss for dense object detection [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020, 42(2): 318-327.
[11] 曹跃腾, 朱学岩, 赵燕东, 等. 基于改进ResNet的植物叶片病虫害识别[J]. 中国农机化学报, 2021, 42(12): 175-181.
Cao Yueteng, Zhu Xueyan, Zhao Yandong, et al. Recognition of plant leaf diseases and insect pests based on improved ResNet [J]. Journal of Chinese Agricultural Mechanization, 2021, 42(12): 175-181.
[12] 宋中山, 汪进, 郑禄, 等. 基于二值化的Faster R—CNN柑橘病虫害识别研究[J]. 中国农机化学报, 2022, 43(6): 150-158.
Song Zhongshan, Wang Jin, Zheng Lu, et al. Research on citrus pest identification based on Binary Faster R—CNN [J]. Journal of Chinese Agricultural Mechanization, 2022, 43(6): 150-158.
[13] 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.
[14] Zhong Y, Zhao M. Research on deep learning in apple leaf disease recognition [J]. Computers and Electronics in Agriculture, 2020, 168: 105146.
[15] Sun H, Xu H, Liu B, et al. Mean—SSD: A novel real‑time detector for apple leaf diseases using improved light‑weight convolutional neural networks [J]. Computers and Electronics in Agriculture, 2021, 189: 106379.
[16] 徐艳蕾, 孔朔琳, 陈清源, 等. 基于Transformer的强泛化苹果叶片病害识别模型[J]. 农业工程学报, 2022, 38(16): 198-206.
Xu Yanlei, Kong Shuolin, Chen Qingyuan, et al. Model for identifying strong generalization apple leaf disease using Transformer [J]. Transactions of the Chinese Society of Agricultural Engineering, 2022, 38(16): 198-206.
[17] Zhu R, Zou H, Li Z, et al. Apple—Net: A model based on improved YOLOv5 to detect the apple leaf diseases [J]. Plants, 2023, 12(1): 169.
[18] Zhu X, Li J, Jia R, et al. Lad—Net: A novel light weight model for early apple leaf pests and diseases classification [J]. IEEE/ACM Transactions on Computational Biology and Bioinformatics, 2023, 20(2): 1156-1169.
[19] Zhang Y, Zhou G, Chen A, et al. A precise apple leaf diseases detection using BCTNet under unconstrained environments [J]. Computers and Electronics in Agriculture, 2023, 212: 108132.
[20] Jiang Y, Tong W. Improved lightweight identification of agricultural diseases based on MobileNetV3 [C]. CAIBDA 2022; 2nd International Conference on Artificial Intelligence, Big Data and Algorithms. VDE, 2022: 1-5.
[21] Ma N, Zhang X, Zheng H T, et al. Shufflenet v2: Practical guidelines for efficient CNN architecture design [C]. Proceedings of the European Conference on Computer Vision (ECCV), 2018: 116-131.
[22] Liu Y, Shao Z, Hoffmann N. Global attention mechanism: Retain information to enhance channel‑spatial interactions [J]. arXiv preprint arXiv: 2112.05561, 2021.
[23] Wang J, Chen K, Xu R, et al. Carafe: Content‑aware reassembly of features [C]. Proceedings of the IEEE/CVF International Conference on Computer Vision, 2019: 3007-3016.
[24] Chen J, Kao S, He H, et al. Run, don't walk: Chasing higher FLOPS for faster neural networks [C]. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023: 12021-12031.
[25] Yang Q, Duan S, Wang L. Efficient identification of apple leaf diseases in the wild using convolutional neural networks [J]. Agronomy, 2022, 12(11): 2784.
[26] Selvaraju RR, Cogswell M, Das A, et al. Grad—CAM: Visual explanations from deep networks via gradient‑based localization [J]. International Journal of Computer Vision, 2020, 128(2): 336.

[1]	Liu Yuping, Liu Chengfei, Zhao Pingwei. Detection method of rice leaf disease based on DeepLabv3—Faster R—CNN [J]. Journal of Chinese Agricultural Mechanization, 2025, 46(4): 108-113.
[2]	Lü Zongwang, , Wang Tiantian, , Sun Fuyan, , Zhu Yuhua, . Improved YOLOv8m-based grain insect detection method for wheat storage [J]. Journal of Chinese Agricultural Mechanization, 2025, 46(3): 108-114.
[3]	Guo Jingtao, Lü Feng, Zhang Huiting, Yang Biao, Liu Dayang. Strawberry target detection method based on transfer learning and lightweight YOLOv5s [J]. Journal of Chinese Agricultural Mechanization, 2025, 46(3): 253-260.
[4]	Gu Rui, , Gu Jiale, Song Cuiling, , Qian Chunhua. Recognition model of multi‑scale lightweight apple leaf disease based on coordinate attention [J]. Journal of Chinese Agricultural Mechanization, 2025, 46(2): 173-180.
[5]	Dai Min, Sun Wenjing, Miao Hong. Identification of pepper pests and diseases based on lightweight CBAM—GoogLeNet [J]. Journal of Chinese Agricultural Mechanization, 2025, 46(2): 224-229.
[6]	Wu Yanghua, Wang Jiannan, Liu Minji, You Zhaoyan, Xie Huanxiong, Du Yuanjie. Real‑time peanut pods detection method based on improved YOLOv5n [J]. Journal of Chinese Agricultural Mechanization, 2025, 46(2): 230-236.
[7]	Wang Xinmiao, Zhang Zheng, Dong Xiaowei, Wang Linfeng, Li Ruixiang. Research on weed detection in millet field based on improved YOLOv8 algorithm [J]. Journal of Chinese Agricultural Mechanization, 2025, 46(1): 185-189.
[8]	Song Lihang, Zhang Yi, Shi Yinhao. Maturity detection method of nectarine in natural environment based on improved YOLOv5s [J]. Journal of Chinese Agricultural Mechanization, 2024, 45(9): 250-257.
[9]	Yang Qiaomei, Cui Tingting, Yuan Yongbang, Luo Hua. Research on the application of lightweight YOLO model in detection of small crop diseases and pests [J]. Journal of Chinese Agricultural Mechanization, 2024, 45(9): 265-270.
[10]	Hu Guoyu, , Liu Guang, Zhou Xingguang, Dong Yalan, Zhou Jianping, . Detection method of grape disease in orchard environment based on Swin-TDL algorithm [J]. Journal of Chinese Agricultural Mechanization, 2024, 45(8): 234-239.
[11]	Feng Xiao, , Zhang Hui, , Liu Yunchao, Zhang Wei, , Li Xiaohong, , Ma Zhongjie, . Detection of grape cluster in stone hardening stage based on improved YOLOv5s [J]. Journal of Chinese Agricultural Mechanization, 2024, 45(8): 240-245.
[12]	Wang Ketao, Chen Shifeng, Chen Gui, Wei Jin, Meng Liwen, Chen Quancheng. Detection method of garden spherical hedge based on YOLOv5s [J]. Journal of Chinese Agricultural Mechanization, 2024, 45(8): 262-269.
[13]	Li Xianna, Wu Qiang, Zhang Yidan, Zhou Kang. Detection of tomato leaf disease in small sample under selfsupervised learning [J]. Journal of Chinese Agricultural Mechanization, 2024, 45(7): 172-179.
[14]	Huang Xiaoyu, Zhang Cong, Chen Xiaoling. Research on image recognition of shaded tomato diseases based on multiscale feature fusion [J]. Journal of Chinese Agricultural Mechanization, 2024, 45(7): 194-200.
[15]	Huang Yao, He Jing, Fu Rao, Liu Gang, , Lin Yuanyang. Application analysis of the YOLOv5s-CBAM algorithm for the identification of eggs of Pomacea [J]. Journal of Chinese Agricultural Mechanization, 2024, 45(6): 223-228.