摘要
针对钢带缺陷传统的人工检测效率低、误检率高以及危险程度大等问题,提出了一种基于机器视觉的缺陷检测和识别的研究方案。采用工业摄像头采集钢带生产线上的视频图像,通过中值滤波和小波分析相结合的方法去噪,并用Canny算子实现边缘检测,再以缺陷图像的圆形度等特征完成识别分类,从而实现对钢带缺陷的检测和统计。实验结果表明,该缺陷检测方案能够实时准确有效地识别钢带缺陷,证明了该方法的可行性。
Aiming at low manual detection efficiency, high rate of false detection and high labor intensity, an algorithm based on machine vision was proposed to achieve automated defect detection and identification of strip steel. Using industrial cameras capture video images from strip steel production line, the method of median filtering and wavelet analysis were combined to eliminate noises; utilizing Canny operator to realize Edge detection, and then completed the identification and classification according to the features of defect images such as circular degree. The experimental result shows that this detection system can accurately and effectively identify the defects in a real time, and proves the feasibility of this method.
出处
《微型机与应用》
2015年第24期50-52,共3页
Microcomputer & Its Applications
关键词
机器视觉
缺陷检测
识别分类
钢带
machine vision
defect defection
identification
strip steel
