国内外板带材表面质量检测方法研究进展

Research progress of surface finishing test methods of board strip at domestic and overseas

随着板带材越来越广泛的应用于工业生产,板带材的表面质量问题也越来越受到用户的关注。本文介绍了传统检测法、自动检测法以及机器视觉检测法三种常用的表面质量检测法,对其中的优缺点进行分析表明,采用面阵CCD技术和线阵CCD技术,结合自学习神经网络分类的机器视觉检测,在生产中应用效果较好。最后总结了板带材检测方法的发展方向:研究更有效的分类识别算法分类识别板带材的表面缺陷;优化图像采集和处理的算法,缩短计算时间。

Board strip is being more and more widely used in industrial production. Its surface quality attracts more and more attention as well. Traditional detection, automation detection and machine vision method are 3 most common surface quality methods. The area CCD technology and a linear CCD technology, combined with self- learning neural network classifier, perform better in the merits and drawbacks production comparison. More effective classification and recognition algorithm analyzes surface defect, optimizing algorithm for image acquisition and processing, shorten the calculation time are development direction of board strip detection.