Detection Algorithm of Switch Machine Gap Based on Template Matching

In view of the shortcomings of the existing machine vision-based automatic detection methods for switch machine gap, this paper proposed a simple gap detection algorithm based on template matching. According to the gap features in the image, a variable scale template was designed combining the gradient direction and size information, where adaptive threshold segmentation and morphological filtering were applied to the acquired gap image, to deform the binary image to enhance the robustness of the matching template and simplify the similarity measurement criterion. The designed variable scale template and similarity measurement were used to detect the gap. Finally, in order to solve the gap boundary occlusion problem due to excessive gap displacement, this paper designed a connected domain extraction scheme to compensate the matching detection results. The experimental results show that the algorithm, simple and fast, with strong robustness to installation angle, distance and illumination variation, has a wide measurement range, with a measurement accuracy of 0. 1 mm in the measurement range of 0-4. 0 mm. © 2021, Department of Journal of the China Railway Society. All right reserved.