An algorithm for extracting spray trajectory based on laser vision

Abstract

Aiming at the need of digital shoe production, a method of automatically generating the spraying trajectory of the shoe is proposed. This method used the 3D sensor to obtain the original three-dimensional point cloud data of the sole, and used the median filtering method to process the data of 3D point data. The method of transforming the spatial trajectory of the soles of the shoe into a two-dimensional image depth map is proposed, and the point P (x, y, z) in the soles of the sole is mapped in the RGB channel of the image pixel. The method of computer vision image processing is used. In this paper, the Sobel, Canny and Laplace algorithm operators are used to analyze and implement the soles of the soles, and the edge contours of the soles are obtained by comparing with the proposed two-block algorithm. The edge contours extracted by the two-plug algorithm are better Then, the B-spline curve fitting algorithm is used to get the offset curve of the edge of the shoe, and finally get the actual robot spray trajectory curve.