Laser curve extraction of train wheelset based on U-Net

The method for rapid and accurate extraction of fringe images of wheel pairs is studied. The classic U-Net network model is used to achieve precise segmentation of laser stripes, and the gray center of gravity method is used to achieve sub-pixel extraction of the segmented image in the form of a template. Firstly, the U-net network model is used to do laser stripe segmentation, then the template method is used to find the center of the light bar, and finally the gray center of gravity method is used to achieve fast and accurate laser curve extraction. Experimental results show that this method can effectively overcome the effects of background noise and bright spots on laser stripe extraction under dynamic environment.