基于机器视觉的工件识别和定位系统设计

Design of Workpiece Recognition and Positioning System Based on Machine Vision

目前在制造工业产线上的机械手大多采用人工输入指令控制执行机构运动,只能完成固定模板下的操作,对于一些复杂工况很可能致使机器人运作失败。这一定程度上导致生产成本升高,工作效率低和自动化程度降低。在制造产线中增加机器视觉技术,将机器视觉中的远程检测、图像识别、定位和引导等技术引用到机器控制系统中,使得机器能够自主应对环境变化,可以提高机器工作的智能化和自动化。本文将机器视觉引入工业机器自动化生产线并通过视觉识别与引导实现工件定位和抓取。着重对系统搭建的关键步骤、相机与机器坐标的统一和工件识别算法进行设计和开发。

At present,most of the manipulators on the manufacturing industrial production line use manual input instructions to control the movement of the actuator,and can only complete the operation under the fixed template,and it is likely to cause the robot operation to fail for some complex working conditions.To some extent,this leads to higher production costs,low work efficiency and reduced automation.Add machine vision technology to the manufacturing line,and introduce remote detection,image recognition,positioning and guidance technologies in machine vision to the machine control system,so that the machine can independently respond to environmental changes,and can improve the intelligence and automation of machine work.This paper introduces machine vision into the automatic production line of industrial machines and realizes workpiece positioning and grasping through visual recognition and guidance.Focus on the key steps of system construction,the unification of camera and machine coordinates,and the design and development of workpiece recognition algorithms.