摘要
利用计算机视觉实现轮胎X光图像的质量缺陷检测,是当前轮胎生产质量监控自动化的关键环节。通过对大量正常轮胎图像纹理特征的研究,发现其X-光图像纹理呈现明显的准周期性和规则性,鉴于此设计了能反映该规律性的小波函数,实现对轮胎内存在异物和钢丝帘线分布不匀等异常检测。实验结果与传统的灰度共生矩阵法相比,本文方法具有对缺陷位置定位精确、运算快速等优点。
Tyro defects dtection by means of computer vision based on X-ray image is the key step in tyro automatlc production. It is found that normal tyres are of regular para-perlodlcal textures by observing large otmt of normal X-ray images, thefore a wavelet function is designed to model this regularity, which can effectively detect fureign matter existence and irregularity of metal curtain distribution. Experiment results prove this method is more accm＇atc and faster than
出处
《信息技术与信息化》
2013年第6期101-103,共3页
Information Technology and Informatization
