基于改进YOLOv5的田间复杂环境障碍物检测

doi:10.13733/j.jcam.issn.2095-5553.2024.06.032

摘要/Abstract

摘要：

为实现田间复杂环境下农业机器人自主导航作业过程中障碍物快速检测，提出一种基于改进YOLOv5的田间复杂环境下障碍物检测方法。建立包含农业机械、人、羊三类目标障碍物共计6 766张图片的农田障碍物数据集；通过k-means聚类算法生成最佳先验锚框尺寸；引入CBAM卷积块注意力模块，抑制目标障碍物周围复杂环境的干扰，增强目标显著度；增加一个检测头，跨层级连接主干特征，增强多尺度特征表达能力，缓解标注对象尺度方差带来的负面影响；使用Ghost卷积替换Neck层中普通卷积，减少模型参数，降低模型复杂度。改进后的模型比YOLOv5s基准模型检测精度提高2.3%，召回率提高3.1%，精确率提高1.9%，参数量降低7%左右，为无人农业机械自主作业过程中导航避障的研究提供技术参考。

关键词: 农业机器人, 农田障碍物检测, 改进YOLOv5, 图像处理, 机器视觉

Abstract:

In order to realize the rapid detection of obstacles in the process of autonomous navigation of agricultural robots in complex field environments, an obstacle detection method based on improved YOLOv5 in complex field environments is proposed. The farmland obstacle data set containing a total of 6766 images of agricultural machinery, human and sheep objects are established. The optimal prior anchor box size is generated by the kmeans clustering algorithm. The CBAM convolution block attention module is introduced to suppress the interference of the complex environment around the target obstacle and enhance the target saliency. A detection head is added to connect the backbone features across levels, enhance the ability to express multiscale features, and alleviate the negative impact of the variance of the scale of the labeled objects. The Ghost convolution is used to replace the ordinary convolution in the Neck layer to reduce the model parameters and decrease the model complexity. Compared with the YOLOv5s benchmark model, the improved model has increased the detection accuracy by 2.3%, the recall rate by 3.1%, the accuracy rate by 1.9%, and has decreased the reference number by about 7%. It provides technical reference for the research of navigation and obstacle avoidance during autonomous operation of unmanned agricultural machinery.

Key words: agricultural robot, farmland obstacle detection, improved YOLOv5, image processing, machine vision

中图分类号:

S220
TP391.4

杨昊霖, 王其欢, 李华彪, 耿端阳, 武继达, 姚艳春. 基于改进YOLOv5的田间复杂环境障碍物检测[J]. 中国农机化学报, 2024, 45(6): 216-222.

Yang Haolin, Wang Qihuan, Li Huabiao, Geng Duanyang, Wu Jida, Yao Yanchuan. Obstacle detection in complex farmland environment based on improved YOLOv5[J]. Journal of Chinese Agricultural Mechanization, 2024, 45(6): 216-222.

参考文献

［1］　薛金林, 李雨晴, 曹梓建. 基于深度学习的模糊农田图像中的障碍物检测技术［J］. 农业机械学报, 2022, 53(3): 234-242.
Xue Jinlin, Li Yuqing, Cao Zijian. Obstacle detection based on deep learning for blurred farmland images ［J］. Transactions of the Chinese Society for Agricultural Machinery, 2022, 53(3): 234-242.
［2］　Mwalupaso G E, Wang S, Rahman S, et al. Agricultural informatization and technical efficiency in maize production in Zambia ［J］. Sustainability, 2019, 11(8): 2451.
［3］　何勇, 蒋浩, 方慧, 等. 车辆智能障碍物检测方法及其农业应用研究进展［J］. 农业工程学报, 2018, 34(9): 21-32.
He Yong, Jiang Hao, Fang Hui, et al. Research progress of intelligent obstacle detection methods of vehicles and their application on agriculture ［J］. Transactions of the Chinese Society of Agricultural Engineering, 2018, 34(9): 21-32.
［4］　郝帅, 杨磊, 马旭, 等. 基于注意力机制与跨尺度特征融合的YOLOv5输电线路故障检测［J］. 中国电机工程学报, 2023(6): 2319-2330.
Hao Shuai, Yang Lei, Ma Xu, et al. YOLOv5 transmission line fault detection based on attention mechanism and crossscale feature fusion ［J］. Proceedings of the CSEE, 2023(6): 2319-2330.
［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］. Proceedings of the IEEE International Conference on Computer Vision, 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］. Advances in Neural Information Processing Systems, 2015, 28.
［8］　夏成楷. 基于深度学习的农田障碍物的识别和无人农业车辆避障策略研究［D］. 南京: 南京农业大学, 2020.
Xia Chengkai. Research on farmland obstacle recognition based on deep learning and obstacle avoidance strategies for unmanned agricultural vehicles ［D］. Nanjing: Nanjing Agricultural University, 2020.
［9］　Liu W, Anguelov D, Erhan D, et al. Ssd: Single shot multibox detector ［C］. Computer VisionECCV, 2016: 14th European Conference, 2016: 21-37.
［10］　Lin T Y, Goyal P, Girshick R, et al. Focal loss for dense object detection ［C］. Proceedings of the IEEE International Conference on Computer Vision, 2017: 2980-2988.
［11］　Redmon J, Farhadi A. YOLO9000: Better, faster, stronger ［C］. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2017: 7263-7271.
［12］　Redmon J, Farhadi A. Yolov3: An incremental improvement ［J］. ArXiv preprint ArXiv: 1804.02767, 2018.
［13］　Bochkovskiy A, Wang C Y, Liao H Y M. Yolov4: Optimal speed and accuracy of object detection ［J］. arXiv preprint arXiv: 2004.10934, 2020.
［14］　刘俊明, 孟卫华. 基于深度学习的单阶段目标检测算法研究综述［J］. 航空兵器, 2020, 27(3): 44-53.
Liu Junming, Meng Weihua. Review on singlestage object detection algorithm based on deep learning ［J］. Aero Weaponry, 2020, 27(3): 44-53.
［15］　赵奇慧, 刘艳洋, 项炎平. 基于深度学习的单阶段车辆检测算法综述［J］. 计算机应用, 2020, 40(S2): 30-36.

Zhao Qihui, Liu Yanyang, Xiang Yanping. Review of onestage vehicle detection algorithms based on deep learning ［J］. Journal of Computer Applications, 2020, 40(S2): 30-36.

［16］　刘慧, 张礼帅, 沈跃, 等. 基于改进SSD的果园行人实时检测方法［J］. 农业机械学报, 2019, 50(4): 29-35, 101.

Liu Hui, Zhang Lishuai, Shen Yue, et al. Realtime pedestrian detection in orchard based on improved SSD ［J］. Transactions of the Chinese Society for Agricultural Machinery, 2019, 50(4): 29-35, 101.

［17］　魏建胜, 潘树国, 田光兆, 等. 农业车辆双目视觉障碍物感知系统设计与试验［J］. 农业工程学报, 2021, 37(9): 55-63.

Wei Jiansheng, Pan Shuguo, Tian Guangzhao, et al. Design and experiments of the binocular visual obstacle perception system for agricultural vehicles ［J］. Transactions of the Chinese Society of Agricultural Engineering, 2021, 37(9): 55-63.

［18］　李文涛, 张岩, 莫锦秋, 等. 基于改进YOLOv3-tiny的田间行人与农机障碍物检测［J］. 农业机械学报, 2020, 51(S1): 1-8, 33.

Li Wentao, Zhang Yan, Mo Jinqiu, et al. Detection of pedestrian and agricultural vehicles in field based on improved YOLOv3-tiny ［J］. Transactions of the Chinese Society for Agricultural Machinery, 2020, 51(S1): 1-8, 33.

［19］　Wang J, Chen Y, Gao M, et al. Improved YOLOv5 network for realtime multiscale traffic sign detection ［J］. arXiv preprint arXiv: 2112.08782, 2021.

［20］　杨娟娟, 高晓阳, 李红岭, 等. 基于机器视觉的无人机避障系统研究［J］. 中国农机化学报, 2020, 41(2): 155-160.

Yang Juanjuan, Gao Xiaoyang, Li Hongling, et al. Research on UAV obstacle avoidance system based on machine vision ［J］. Journal of Chinese Agricultural Mechanization, 2020, 41(2): 155-160.

［21］　刘路, 潘艳娟, 陈志健, 等. 高遮挡环境下玉米植保机器人作物行间导航研究［J］. 农业机械学报, 2020, 51(10): 11-17.

Liu Lu, Pan Yanjuan, Chen Zhijian, et al. Interrows navigation method for corn crop protection vehicles under high occlusion environment ［J］. Transactions of the Chinese Society for Agricultural Machinery, 2020, 51(10): 11-17.

［22］　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.

［23］　Han K, Wang Y, Tian Q, et al. GhostNet: More features from cheap operations ［C］. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020: 1580-1589.

[1]	王雷, 钟康生, 郭小宝, 盛斌, 肖波, 荆磊. 基于机器视觉的陈皮热泵干燥含水率预测模型研究[J]. 中国农机化学报, 2024, 45(7): 97-103.
[2]	张延军, 赵建鑫, . 基于多路神经网络的黄花菜关节点检测算法研究[J]. 中国农机化学报, 2024, 45(7): 228-234.
[3]	陈泉淦, 陈新元, 曾镛, 程志文. 基于YOLOv5的机耕船双目视觉障碍感知研究[J]. 中国农机化学报, 2024, 45(7): 261-268.
[4]	郎秀丹, 史宇亮, 李天华, 陈明东. 基于阈值识别的温室黄瓜整体自动落蔓装置研究[J]. 中国农机化学报, 2024, 45(6): 106-112.
[5]	邱畅, 田光兆, 赵嘉威, 谢尚杰, 郑奎. 基于改进YOLOv5的草莓病害识别[J]. 中国农机化学报, 2024, 45(3): 198-204.
[6]	缪亚伦, 石美琦, 孟海涛, 梁旭升, 黄才贵, 李岩舟. 基于Stereo CameraYOLOv5自然环境下百香果检测与定位模型[J]. 中国农机化学报, 2024, 45(3): 233-241.
[7]	张伟进, 王福顺, , 孙小华, 王军皓, 刘宏权, 王鑫鑫, . 传统图像分割算法在农作物籽粒考种应用中的研究进展[J]. 中国农机化学报, 2024, 45(2): 280-287.
[8]	王焱清, , 汤旸, 杨光友, . 面向机器人柑橘采摘的控制系统设计与试验[J]. 中国农机化学报, 2023, 44(9): 146-153.
[9]	侯义锋, , 钱俊, , 王梁, 贺杰, 闭耀锋. 砂糖橘采摘机器人分拣机构设计[J]. 中国农机化学报, 2023, 44(9): 183-189.
[10]	黄家才, 汪涛, 张铎, 唐安, 高芳征. 基于改进YOLOv4的草莓果柄叶识别与定位[J]. 中国农机化学报, 2023, 44(9): 205-213.
[11]	王大明, 何逸霏, 李华英, 苟于江, 何辉波. 柑橘采摘机器人图像识别算法研究[J]. 中国农机化学报, 2023, 44(9): 222-226.
[12]	姜来, 王文娣, 霍晓静, 王辉, 唐娟, 李丽华, (. 死鸡识别机器人系统设计与试验[J]. 中国农机化学报, 2023, 44(8): 81-87.
[13]	刘浩, 孟令峰, 王松峰, 刘自畅, 杜海娜, 孙福山. 基于机器视觉的烤烟鲜烟成熟度判别模型优选[J]. 中国农机化学报, 2023, 44(8): 118-124.
[14]	舒卓, 陈立平, 陈梅香, 张瑞瑞, 郭新宇, 温维亮. 单目多视角图像的美国白蛾三维模型重建系统与试验[J]. 中国农机化学报, 2023, 44(8): 168-173.
[15]	李航, 廖映华, 黄波. 基于改进DQN算法的茶叶采摘机械手路径规划[J]. 中国农机化学报, 2023, 44(8): 198-205.