数学形态学梯度算子的钢轨端面几何特征测量

Measurement of rail end surface geometrical characteristics based on mathematical morphology gradient algorithm

针对传统的接触式钢轨端面几何特征测量误差大且测量结果不稳定的问题,基于数学形态学梯度算子的方法实现钢轨端面几何特征的非接触测量.测量采用工业CCD数码相机采集钢轨端面图像数据,将其灰度处理并得到图像的二值矩阵,用数学形态学运算获得钢轨端面图像的边缘,结合Sobel梯度算子实现钢轨端面几何尺寸的测量.测量数据显示,该方法可获得比较准确的钢轨端面几何特征测量值.

Aiming at the problems of poor measurement precision and unstable measurement data in the traditional measurement method on the rail end surface geometrical characteristics, non-touching measurement method is proposed based on mathematical morphology gradient algorithm. The industry CCD digital camera is used to acquire the image data ofthe rail end surface, the colorized image of the rail end surface is converted into values of gray image and binary matrix, the image edge of the rail can be acquired by the mathematical morphology operation with the binary image matrix, and the geometrical dimensions of the rail end surface can be measured from the image edge based on the Sobel gradient algorithm. The measurement results show that the novel method can acquire accurate measurement values of geometrical characteristics on the rail end surface.