基于SE-MultiResNet50算法的辣椒病害种类及程度分级检测

doi:10.13733/j.jcam.issn.20955553.2024.12.038

摘要/Abstract

摘要：

在实际的辣椒种植环境中，由于其复杂背景，辣椒叶片病害的识别难度较大。目前，关于受害程度分级检测和辣椒病害分级缺乏公开的数据集。以成都市农林科学院辣椒种植基地的叶片为研究对象，采用U2-Net对叶片进行分割，生成具有不同复杂背景的合成图像，从而丰富数据集。针对常见的辣椒细菌性斑点、白粉病和病毒病3种病害以及健康叶片，提出一种SE-MultiResNet50检测模型。该模型在全由复杂背景图像组成的测试集上表现出色：辣椒病害种类的识别准确率达到91.05%，受害严重程度分级的准确率为92.08%。结果表明，该检测模型在复杂背景下具有较高的识别精度，成功实现对辣椒病害种类分类和受害严重程度分级的智能识别。同时，提供一种新的数据集扩充方法，为相关领域的研究提供新的思路和途径。

关键词: 辣椒, 病虫害, 分级检测, 注意力机制, ResNet50

Abstract:

In the actual pepper planting environment, due to its complex background, the identification of pepper leaf diseases has always been a challenging problem. Currently, there is a lack of publicly available datasets for severity grading detection and classification of chili pepper leaf diseases. This study focuses on leaf samples from the chili pepper plantation base of the Chengdu Academy of Agriculture and Forestry Sciences, by utilizing U2-Net for leaf segmentation to generate synthetic images with diverse complex backgrounds, thereby enriching the dataset. Addressing common chili pepper diseases such as bacterial spot, powdery mildew, viral diseases, and healthy leaves, a SE-MultiResNet50 detection model is proposed. This model performs remarkably well on a test set comprised entirely of images with complex backgrounds: achieving a recognition accuracy of 91.05% for chili pepper disease types and 92.08% for severity grading. Experimental results of this study demonstrate that the detection model exhibits high recognition accuracy under complex backgrounds, successfully achieving intelligent identification of chili pepper disease types and severity grading. Additionally, a novel dataset augmentation method is provided, offering new insights and avenues for research in related fields.

Key words: chili, diseases and pests, hierarchical detection, attention mechanism, ResNet50

中图分类号:

S436.418

唐源, 鲁茂悦, 李丽平, 唐有万, 陈阳洋, 李昱瑾. 基于SE-MultiResNet50算法的辣椒病害种类及程度分级检测[J]. 中国农机化学报, 2024, 45(12): 259-266.

Tang Yuan, Lu Maoyue, Li Liping, Tang Youwan, Chen Yangyang, Li Yujin. Grading detection of chili pepper diseases species and degree based on the SE-MultiResNet50 algorithm[J]. Journal of Chinese Agricultural Mechanization, 2024, 45(12): 259-266.

参考文献

［1］　贾童童. 植物病虫草害防治中生物技术的应用分析［J］. 科教导刊(电子版), 2018(4): 253-254.
［2］　梁栋, 刘娜, 张东彦, 等. 利用成像高光谱区分冬小麦白粉病与条锈病［J］. 红外与激光工程, 2017, 46(1): 42-50.
Liang Dong, Liu Na, Zhang Dongyan, et al. Discrimination of powdery mildew and yellow rust of winter wheat using highresolution hyperspectra and imageries ［J］. Infrared and Laser Engineering, 2017, 46(1): 42-50.
［3］　马超, 袁涛, 姚鑫锋, 等. 基于HOG+SVM的田间水稻病害图像识别方法研究［J］. 上海农业学报, 2019, 35(5): 131-136.
Ma Chao, Yuan Tao, Yao Xinfeng, et al. Study on image recognition method of rice disease in field based on HOG+SVM ［J］. Acta Agriculturae Shanghai, 2019, 35(5): 131-136.
［4］　Aziz S, Bashir M, Mughal O, et al. Image pattem classification for plant disease identification using local tridirectional features ［C］. 2019 IEEE 10th Annual Information Technology, Electronics and Mobile Communication Conference (IEMCON).IEEE, 2019.
［5］　Jiang P, Chen Y, Liu B, et al. Realtime detection of apple leaf diseases using deep learning approach based on improved convolutional neural networks ［J］. Ieee Access, 2019, 7: 59069-59080.
［6］　Zhang S, Huang W, Zhang C. Threechannel convolutionalneural networks for vegetable leaf disease recognition ［J］. Cognitive Systems Research, 2019, 53: 31-41.
［7］　Arsenovic M, Karanovic M, Sladojevic S, et al. Solving current limitations of deep learning based approaches for plant disease detection ［J］. Symmetry, 2019, 11(7): 939.
［8］　Singh S, Gupta I, Gupta S, et al. Deep learning based automated detection of diseases from apple leaf images ［J］. Computers, Materials & Continua, 2022, 71(1): 1849-1866.
［9］　Liang Q, Xiang S, Hu Y, et al. PD2SENet: Computerassisted plant disease diagnosis and severity estimation network ［J］. Computers and Electronics in Agriculture, 2019, 157: 518-529.
［10］　万军杰, 祁力钧, 卢中奥, 等. 基于迁移学习的 GoogLeNet 果园病虫害识别与分级［J］. 中国农业大学学报, 2021, 26(11): 209-221.
Wan Junjie, Qi Lijun, Lu Zhongao, et al. Recognition and grading of diseases and pests in orchard by GoogLeNet based on transfer learning ［J］. Journal of China Agricultural University, 2021, 26(11): 219-221.
［11］　Pratap V K, Kumar N S. Highprecision multiclass classification of chili leaf disease through customized EffecientNetB4 from chili leaf images ［J］. Smart Agricultural Technology, 2023, 5: 100295.
［12］　李西兴, 陈佳豪, 吴锐, 等. 基于改进MaxViT的辣椒病害识别分类方法［J］. 华中农业大学学报, 2024, 43(2): 123-133.
Li Xixing, Chen Jiahao, Wu Rui, et al. A method for identifying and classifying pepper diseases based on improved MaxViT ［J］. Journal of Huazhong Agricultural University, 2024, 43(2): 123-133.

［13］　Uzhinskiy A V, Ososkov G A, Goncharov P V, et al. One-shot learning with triplet loss for vegetation classification tasks ［J］. Компьютерная оптика, 2021, 45(4): 608-614.

［14］　GB/T 17980.34—2000, 农药田间药效试验准则(一)杀菌剂防治梨黑星病［S］.
［15］　Qin X, Zhang Z, Huang C, et al. U2-Net: Going deeper with nested Ustructure for salient object detection ［J］. Pattern recognition, 2020, 106: 107404.
［16］　Ronneberger O, Fischer P, Brox T. U-Net: Convolutional networks for biomedical image segmentation ［C］. Medical Image Computing and ComputerAssisted InterventionMICCAI 2015: 18th International Conference, Munich, 2015: 234-241.
［17］　Zhang L, Shen Z, Lin W, et al. U2 Netbased singlepixel imaging salient object detection ［J］. Current Optics and Photonics, 2022, 6(5): 463-472.
［18］　Wang L, Lu H, Wang Y, et al. Learning to detect salient objects with imagelevel supervision ［C］. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2017: 136-145.
［19］　He K, Zhang X, Ren S, et al. Deep residual learning for image recognition ［C］. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016: 770-778.
［20］　Ioffe S. Batch normalization: Accelerating deep network training by reducing internal covariate shift ［J］. arxiv preprint arxiv: 1502.03167, 2015.
［21］　Olshausen B A, Anderson C H, Van Essen D C. A neurobiological model of visual attention and invariant pattern recognition based on dynamic routing of information ［J］. Journal of Neuroscience, 1993, 13(11): 4700-4719.
［22］　Hu J, Shen L, Sun G. Squeezeandexcitation networks ［C］. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2018: 7132-7141.
［23］　Jaderberg M, Simonyan K, Zisserman A.Spatial transformer networks ［J］. Advances in Neural Information Processing Systems, 2015, 28.
［24］　Woo S, Park J, Lee J Y, et al. Cbam: Convolutional block attention module ［C］. Proceedings of the European Conference on Computer Vision (ECCV), 2018: 3-19.

[1]	李新澳, 刘元义, 王丽娟, 张策, 陶丽欢, 张悦. 设施农业臭氧杀菌闭环控制系统设计与试验[J]. 中国农机化学报, 2024, 45(9): 62-68.
[2]	韩鹏飞, 宋其江, 贾梦实. 基于改进轻量化 EfficientNet-V2模型的小麦种子分类[J]. 中国农机化学报, 2024, 45(9): 111-117.
[3]	付晓鸽, 李涵, 左治江, 杜铮 . 基于 Mask R-CNN的复杂环境下辣椒识别方法研究[J]. 中国农机化学报, 2024, 45(9): 215-219.
[4]	张传栋, 亓璐, 丁华立. 基于改进 YOLOv8n模型的多品种葡萄簇检测方法[J]. 中国农机化学报, 2024, 45(9): 220-226.
[5]	杨巧梅, 崔婷婷, 袁永榜, 罗桦 . 轻量化 YOLO模型在农作物微小病虫害检测中的应用研究[J]. 中国农机化学报, 2024, 45(9): 265-270.
[6]	李洁, 高尚兵, 余骥远, 陈新, 李士丛, 袁星星. 基于PMMS-Net和叶绿素荧光成像的绿豆叶斑病抗性鉴定方法[J]. 中国农机化学报, 2024, 45(8): 210-216.
[7]	刘星, 顾寄南, 黄则栋, 张文浩, 张伟. 基于深度学习的苹果点云语义分割方法研究[J]. 中国农机化学报, 2024, 45(8): 223-227.
[8]	苏斐, 王光辉, 史艳霞, 贾然, 闫银发, 祖林禄, . 融合通道注意力机制的ResUnet作物病斑分割模型[J]. 中国农机化学报, 2024, 45(8): 228-233.
[9]	冯晓, 张辉, 刘运超, 张微, 李小红, 马中杰, . 基于改进YOLOv5s的硬核期葡萄簇检测[J]. 中国农机化学报, 2024, 45(8): 240-245.
[10]	周思捷, 刘天奇, 陈天华. 基于改进YOLOv5算法的水稻病害识别研究[J]. 中国农机化学报, 2024, 45(8): 246-253.
[11]	王克涛, 陈世锋, 陈贵, 韦锦, 蒙丽雯, 陈泉成. 基于YOLOv5s的园林球形绿篱检测方法[J]. 中国农机化学报, 2024, 45(8): 262-269.
[12]	岳耀华, 张伟, 亓立强, . 基于改进YOLOv5的田间大豆花朵生长状态检测方法研究[J]. 中国农机化学报, 2024, 45(7): 188-193.
[13]	吴利刚, 吕媛媛, 周倩, 陈乐, 张梁, 史建华. 基于改进YOLOv5的轻量级黄花成熟检测方法[J]. 中国农机化学报, 2024, 45(7): 235-242.
[14]	熊志刚, 陈为真. 一种YOLOv5-DK的柠檬初期病虫害检测算法研究[J]. 中国农机化学报, 2024, 45(7): 249-254.
[15]	张书贵, 陈书理, 赵展. 改进YOLOv8的农作物叶片病虫害识别算法[J]. 中国农机化学报, 2024, 45(7): 255-260.