Abstract:Aiming at the background of incomplete cone workpiece model fitting problem,in order to solve the standard geometric model fitting problem of a large number of point cloud data,a fitting algorithm for cone point cloud data based on genetic algorithm was proposed:by randomly selecting the fitting model sampling points,the distance calculation of point cloud data points-standard model sampling points was used to replace the distance calculation of point cloud data points-standard surface model,so as to improve the calculation efficiency and retain the detailed characteristics of point cloud;by establishing Kd-tree index on point cloud data,the distance detection efficiency of sampling points of standard model was improved;the parameters of the fitting model were searched iteratively by genetic algorithm,and the search efficiency of genetic algorithm was improved by changing the amount of data and fitting accuracy.The experimental results show that the algorithm can effectively improve the matching accuracy of some conical workpiece fitting,with an error range of ±0.5 mm,and has a good effect on the processing of a large number of point cloud data.