搬运机器人多目标识别系统的研究

A Multiple Target Recognition System of Work Piece Handling Robots

机器视觉与工业机器人相结合,构建用于多种目标识别的搬运机器人系统实验平台。在完成机器人坐标系与相机坐标系转换的基础上,以4种不同形状的等高工件为对象,通过CCD相机连续不断地获取运动中的工件图像信息,对采集到的图像进行处理,以识别目标工件、提取工件中心坐标点,然后将工件中心坐标信息传输给机器人,最终实现机器人对工件的定点抓取。实验表明,该系统具备较好的目标识别和定位效果,有助于改善流水线上输送工件单一的状况,提高生产效率。

An experimental platform of multiple target recognition system of work piece handling robots based on machine vision and industrial robot was constructed.On the basis of the complete robot coordinate system and camera coordinate system,the paper studies 4work pieces of the same height,acquired the work piece images using CCD camera,and then processed the images to achieve the recognition of target artifacts,extract the center coordinate points,and finally transferred the central coordinates of the work pieces to the robot.Experiment results show that the system can obtain better target recognition and location effect,which can improve the single condition of the transmission line and improve the production efficiency.