摘要
以MOTOMAN-UP6机器人为基础,构建了一个基于机器视觉的机器人分拣实验系统。该分拣实验系统由机器人、PC机、相机、图像采集卡、传送带和自开发的分拣控制软件组成。系统的工作原理和工作过程为:当目标对象源源不断地进入分拣作业区时,通过相机连续自动地获取作业对象图像,然后由软件对采集到的图像进行运算分析、变换目标对象坐标、识别目标对象分类信息、维护分拣目标的运动踪迹,最终控制机器人实现分拣动作。"单目标跟踪——抓取——搬运"实验和"多目标跟踪——抓取——搬运"实验证明了该技术的意义和实用价值。
Based on MOTOMAN-UP6 robot, a sorting experimental system founded on robot vision is built. This experimental system is composed of robot, PC, camera, image acquisition card, conveyer belt and self-developed sorting control software. The operational principle and process are that targets＇ images are acquired automatically and continuously by the camera, when targets enter the sorting operation area, and these images are dealt by sorting control software with mathematical analysis, transforming the targets＇ coordinates, identifying the targets＇ classification and maintaining their traces, and finally control the robot to implement the sorting action. The experiments of ＂single target tracking-snatching-transit＂ and ＂multi-target tracking- snatching-transit＂ prove the significance and practical value of this technology.
出处
《组合机床与自动化加工技术》
北大核心
2010年第2期84-87,共4页
Modular Machine Tool & Automatic Manufacturing Technique
关键词
工业机器人
分拣
坐标系
目标识别
机器视觉
industrial robot
sorting
coordinate system
target recognition
robot vision
