以注塑制品的四种常见表面缺陷(油污、点缺陷、压痕、划痕)为研究对象,设计了基于机器视觉技术和图像处理技术的自动识别系统,对注塑制品表面缺陷进行检测.介绍了软件系统结构及人机界面设计,并对图像处理的实验结果进行了讨论,制品缺陷的平均识别率达84.44％.系统在解决传统检测方法难题的同时,提高了注塑制品表面缺陷识别的准确性和自动化程度.%Four type surface defects of plastic products ( oil stain, point defects, dents, scratches ) were studied. An automatic recognition system based on machine vision and image processing technology were designed to find surface defects of plastic products.The structure of the software system and the design of human-computer interface were introduced.By analyzing the results of image processing, the average recognition rate could reach 84.44%.Not only the problem of traditional recongnition methods could be solved, but also the accuracy of recognition system to surface defects of plastic products and the degree of automation could be improved.
展开▼
