人机协作视觉手部保护系统设计

针对人机协作中人与机器人共享工作空间时的安全问题,设计了一套人机协作视觉手部保护系统,并搭建相应的验证系统。该系统采用深度学习目标检测算法结合双目视觉技术实现对操作人员手部的识别与定位,同时利用手眼标定将视觉定位后的手部坐标转换到机器人基座坐标系下,通过计算操作人员手部与机器人末端执行器之间的距离,机器人自主执行减速、急停等安全策略。经实验验证:当操作人员在机器人工作空间作业时,通过检测手-末端相对位置关系,可以有效避免人机协作过程中机器人末端执行器与手部发生碰撞,达到了保护操作人员安全的目的。

Design of Visual Hand Protection System for Human-Robot Collaboration

Aiming at the safety problem of operators and robots sharing the working space in human-robot collaboration,this paper designs a set of visual hand protection system for human-robot collaboration.The system uses a deep learning object detection algorithm combined with binocular vision technology to realize the recognition and positioning of the operator's hand,and hand-eye calibration is used to convert the hand coordinate to the robot base coordinate.By calculating the distance between the operator's hand and the robot's end effector,the robot autonomously executes safety strategies such as deceleration and emergency stop.Experiment show that when the operator is working in the robot workspace,by detecting the relative position of the hand and the robot's end effector,it can effectively avoid collision during the human-robot collaboration process,achieving the purpose of protecting the safety of operators.