基于机器视觉的陶瓷瓦表面鼓包缺陷检测算法研究  

RESEARCH ON DEFECT DETECTION ALGORITHM OF CERAMIC TILE SURFACE BULGING DEFECTS BASED ON MACHINE VISION

作　　者：张绍伟 曾曙光[1] 郑胜[1] 肖焱山[1] 李小磊 李强 ZHANG Shao-wei;ZENG Shu-guang;ZHENG Sheng;XIAO Yan-shan;LI Xiao-lei;LI Qiang(College of Science, China Three Gorges University, Yichang 443002, China)

机构地区：[1]三峡大学理学院

出　　处：《南阳理工学院学报》2019年第2期36-41,共6页Journal of Nanyang Institute of Technology

基　　金：国家自然科学基金(U133113)

摘　　要：针对陶瓷瓦表面鼓包缺陷自动化检测的需要,本文提出了基于机器视觉的表面鼓包缺陷检测算法。首先,对陶瓷瓦图像进行预处理,降低噪声。其次,通过自定义滑动滤波、线性中值滤波、插值低通滤波处理来提高鼓包与背景对比度,再采用阈值分割方法及形态学方法将鼓包区域提取出来。最后,通过特征提取得到缺陷信息。实验结果表明,该算法可以实现陶瓷瓦复杂表面的鼓包缺陷的检测,准确率达93%,能够将其应用于陶瓷瓦表面鼓包缺陷检测的生产实践中。Aiming at the need of automatic detection of surface drum defects of ceramic tile, this paper proposes a surface bulge defect detection algorithm based on machine vision. First, the ceramic tile image is preprocessed to reduce noise. Secondly, the drum package and background contrast are improved by custom sliding filter, linear median filtering and interpolation low-pass filtering. Then, the blasting area is extracted by threshold segmentation method and morphological method. Finally, defect information is obtained by feature extraction. The experimental results show that the algorithm can detect the bulge defects of ceramic tile complex surface with an accuracy of 93%, which can be applied to the production practice of ceramic drum surface bulge defect detection.

关 键 词：陶瓷瓦 机器视觉 鼓包检测 滑动滤波 特征提取 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]