Abstract: Focusing on the problem of low accuracy of the rail identification in the existing rail detection algorithm, this paper proposes an algorithm based on Hough transform and Catmull-Rom spline curve. Firstly, the image is denoised and pre-processed, and Canny edge detection algorithm is used for edge detection. Then Hough transform is used to detect the straight track, and the track model is determined according to the vanishing point of the straight track. Finally, the curved track is searched for the characteristic points of the curved track using the bifurcation window. The selected feature points are fitted with Catmull-Rom spline curve fitting curves. Experiments show that the algorithm not only can accurately fit the rail curve, but also displays good accuracy and robustness.

Key words: rail inspection, curved rail fitting, Catmull-Rom spline curve