A New K-Nearest Neighbors Search Algorithm Based on 3D Cell Grids

Based on 3D cell grids,we proposed a new K-nearest neighbors search algorithm.The point cloud was divided twice and distributed to 3D cell grids,then ultimate space,internal space,external space were decided for each grid.With the help of each point's sphere space,K-nearest neighbors of the point can be found quickly.Compared with the existed methods,the proposed algorithm has more efficient performance.