基于YOLOv5的起重机制动下滑量测量系统设计
肖号迪 1吴峰崎 2刘龙 1胡佳伟1
作者信息
- 1. 上海海事大学物流工程学院 上海 200240
- 2. 上海市特种设备监督检验技术研究院 上海 200062
- 折叠
摘要
针对起重机制动下滑量测量问题,现阶段常见检测方法不能满足测量需求,故设计了一种基于YOLOv5的起重机制动下滑量检测软件.文中通过相机采集起重机制动过程,利用YOLOv5s训练并识别标记点,对标记点进行边缘提取和中心定位,改进欧式距离进行图像匹配,使匹配结果更加精确.实验结果表明,改进欧氏距离在匹配上更为精确,明显提高了起重机制动下滑量的测量精度,且下滑量测量系统具有高效性和稳定性.
Abstract
Currently,the common measurement methods of crane braking slip can not meet the measurement requirements.Therefore,in this paper,a crane braking slip detection software based on YOLOv5 is proposed.The braking process of crane was captured by camera,and the marked points were trained and identified by YOLOv5s,so as to extract the edges and locate the center of the marked points,and the Euclidean distance was modified for image matching,which made the matching result more accurate.The experimental results show that the modified Euclidean distance improves the matching accuracy,obviously increases the measurement accuracy of crane braking slip,and the slip measurement system is efficient and stable.
关键词
起重机/制动下滑量/YOLOv5目标检测/特征点匹配/检测软件/设计Key words
crane/brake slip/YOLOv5 target detection/feature point matching/detection software/design引用本文复制引用
基金项目
国家市场监督管理总局科技计划项目(2022MK032)
上海市市场监督管理局科技计划项目(2022-31)
出版年
2024
国家科技期刊平台
NSTL
万方数据