图像处理技术在钢轨表面缺陷检测和分类中的应用

采用图像处理、模式识别及机器视觉等理论,对现役钢轨缺陷进行检测和分类.完成自动提取缺陷图像和最小化缺陷图像,以减少处理量并降低存储空间需求,自动判断缺陷类别.文章对采集到的缺陷图像进行处理,实验结果证明该方法能够正确实现检测轨道表面缺陷检测,并具有一定的适用性.此方法可以克服人工检测方法的许多弊端,提高检测速度和精度.

Application of image processing in defect detection and classification of rail surface

Aiming at the fast development of the railway, an method of defect detection and classification of rail surface was proposed. The new inspection method was based on image processing, pattern recognition and machine vision. Abstracting of defect image, minimizing the processing image to reduce the amount of data handled and the space occupying, classifying defects would be done automatically in this inspection method. The result data from the experiment, which was done with numbers of defect images, showed that this method was capable of classifying the defect images into right group correctly and had stated applicability. The proposed method could overcome the disadvantages of manual inspections and improve the speed and resolution of inspection.