基于Sobel算子的无砟轨道道床板裂缝自动检测算法

Automatic Crack Detection Algorithm for Ballastless Track Slab Based on Sobel Operator

目前无砟轨道道床板裂缝检测采用的电磁、超声等无损检测方法存在检测设备复杂、检测精度低等问题,为此文章提出了基于Sobel算子的无砟轨道道床板裂缝自动检测算法,此算法通过对裂缝图像的预处理、边缘特征提取、边缘标定、长度测量,实现对裂缝方向和长度的判断。实验表明,利用Sobel算子提取无砟轨道道床板裂缝的边缘特征,具有算法简单、裂缝识别准确和对噪声鲁棒的优点,为裂缝检测提供了新的方法。

The cracks of ballastless track slab are detected presently by non-destructive methods such as electromagnetic and ultrasonic,where the problems of complex detection equipment and low detection accuracy exist.Therefore,this paper proposes an automatic crack detection algorithm for the slab based on Sobel operator,with which the direction and length of the cracks can be judged through image preprocessing,edge feature extraction,edge calibration and length measurement.The experimental results show that the Sobel operator is simple,accurate and robust to noise in extracting edge features of the cracks,which provides a new approach for crack detection.