摘要
文章介绍了用于装配线上零件自动识别的方法,选取零件的颜色特征与几何形状特征,将两者组合成有意义的图像特征,作为识别零件的依据,本文针对实际的目标图像,通过对其进行图像预处理,再经过边缘检测求得零件的形状参数,重心坐标和零件的方向,获得零件图像的识别信息。实践表明,该算法简单,速度快,抗干扰能力强,能满足实际使用的要求。
This paper proposes a method of automatic recognizing the image of a work piece, which is designed for the assembly line. This method uses the color feature and geometry shape feture,constituting both meaningful picture characteristic, as gists to recognize component parts.In this paper,the practical object image is inputted and dealt with.Through checking the edge of the image method,shape parameters,the center of gravity coordinates and orientation of the work piece can be picked - up,so we get the recognization information for the component parts.The practice shows the arithmetic is rapid,simpleand robust,and the result of the design accords with the practical request.
出处
《组合机床与自动化加工技术》
2007年第12期58-60,共3页
Modular Machine Tool & Automatic Manufacturing Technique
关键词
机器视觉
颜色特征
形状特征
图像识别
machine vision
color feature
shape feature
image identification
