Research on Point Cloud Reduction Algorithm for Three-dimensional Reconstruction of Prosthesis

Based on the three-dimensional printing of prosthesis limbs, this paper proposes a three-dimensional point cloud reduction algorithm.This algorithm combines the bounding box algorithm and the curvature reduction algorithm.On the basis of not affecting the accuracy of three-dimensional reconstruction, it solves the problem of too many redundant points of point cloud data, accelerates the three-dimensional reconstruction speed and improves the efficiency of prosthesis production.First,the point cloud data is segmented by the bounding box algorithm.Second,the curvature can be calculated, and then the curvature average value can be calculated.Third,according to the mean curvature as the threshold, the flat region and the abrupt region are determined.The bounding box method is used for the flat region, and the curvature reduction method is used for the abrupt region.The new algorithm not only solves the problem of feature point loss caused by the bounding box method, but also solves the local cavity problem caused by the curvature method, improves the efficiency of three-dimensional reconstruction.