摘要
针对机器人带电作业自动化的课题,通过机器视觉方法进行目标物体的定位。分析了像素坐标与世界坐标之间的转换关系,完成相机标定。基于形状匹配算法,提出分块与拼接策略,优化识别效果。基于最优化理论,分析定位误差,提出优化策略。通过实验验证了视觉算法的正确性和可靠性,为机器人自动化作业提供了保障。
In order to solve the automation of the live work robot,machine vision method is used to locate the target object.The conversion relationship between pixel coordinates and global coordinates is analyzed,and the camera calibration is completed.Based on the shape matching algorithm,a block and splicing strategy is proposed to optimize the recognition effect.Based on the optimization theory,the positioning error is analyzed and an optimization strategy is proposed.Experiments have verified the correctness and reliability of the vision algorithm,which provides a guarantee for robot automation.
作者
卞越洋
高晓科
张伟军
BIAN Yueyang;GAO Xiaoke;ZHANG Weijun(School of Mechanical Engineering,Shanghai Jiao Tong University,Shanghai 200240,China)
出处
《机械与电子》
2021年第5期57-61,共5页
Machinery & Electronics
基金
国家重点研发计划资助项目(2017YFC0209902)。
关键词
机器视觉
目标检测
自动化作业
优化策略
machine vision
target detection
automation
optimization strategy
