Research on 3D reconstruction method of heterogeneous workpieces based on line laser scanning

Abstract

In view of the large variation of surface curvature of heterogeneous workpieces and the large individual differences of casting blanks, there will be empty stroke and excessive removal when using robots to process according to the manual teaching path, a method combining robot motion and line laser scanning is proposed to obtain the complex 3D morphology of heterogeneous workpieces, which provides the basic conditions for the research of robot self-adaptive grinding. The robot grabs the line laser calibrated by the hand and eye to scan the heterogeneous workpieces, obtains the 3D point cloud data of the workpiece from different perspectives, filters and simplifies the point cloud, then splices the point cloud image with the coarse registration based on FPFH algorithm and the precise registration based on ICP algorithm, and finally completes the 3D reconstruction of the workpiece surface. The reconstructed model is measured experimentally and compared with the actual size of the workpiece, the size deviation of the model is less than 0.1mm, which shows that this method has high reduction accuracy.