摘要
针对玻璃深加工过程中产生的低对比度表面缺陷,提出了一种基于机器视觉的检测识别算法。首先采用灰度形态学顶帽变换,补偿非均匀照明对图像的影响,其次通过Otsu’s最佳全局阈值分割,获取含缺陷的二值图像,然后设计了一种连通区域分析算法来优化低对比度表面缺陷特征。实验结果表明,该算法可以准确、快速地实现微型平板玻璃低对比度表面缺陷的识别,在一定程度上满足微型平板玻璃表面缺陷自动检测需求。
For low-contrast surface defects of glass emerged in the further processing process,a detection algorithm based on machine vision is proposed.Firstly,the morphological top-hat transform is adopted to compensate the influence of nonuniformillumination on the image.Secondly,Otsuys optimal global threshold segmentation is used to obtain the binary image with defects.In particular,a connected region analysis algorithm is designed to optimize the low-contrast surfacedefect features.The experimental results show that the algorithm can accurately and quickly identify the low-contrast surface defects of micro plate glass.To a certain extent,the requirement of automatic detection of miniature flat glasssurface defects can be met.
作者
李长有
刘遵
李帅涛
LI Changyou;LIU Zun;LI Shuaitao(School of Mechanical and Power Engineering, Henan Polytechnic University, Jiaozuo 454003, China)
出处
《机械工程师》
2018年第3期21-23,27,共4页
Mechanical Engineer
基金
河南省教育厅自然科学基础研究计划项目(2010B510012)
关键词
玻璃表面缺陷
图像处理
机器视觉
缺陷检测
自动化检测
glass surface defect
image processing
machine vision
defect detection
automatic detection
