Kd-tree Based Nonuniform Simplification of 3D Point Cloud

For the over data density of point cloud that greatly affects the model reconstruction efficiency, a nonuniform simplification algorithm for point cloud with normal is presented. At first, kd-tree is used to represent the spatial topology relationships among the point cloud. According to the point density and expectative k-nearest neighbors, the radius of the bounding sphere is calculated to create the sphere centered at the point of the point cloud. Then, the local normal variance and the number of remained points of the neighbors are calculated according to the neighbors of the center point of the sphere, thus determining both their thresholds. The experimental results show that the proposed simplification approach has higher operation efficiency and can avoid holes.