Point Cloud Data Simplification Using Movable Mesh Generation

In order to realize massive point cloud data simplification, a new movable mesh generation algorithm was proposed. Firstly, point cloud model was divided into a number of spatial grids. Secondly, the mesh generation was made like the first step again but more finely and with respecting of the distance threshold, and then, each appropriate point of the grids generated by the second step was filtered out according to the weight values. At last, the position of the minimum point of point cloud model's bounding box was moved, next do the mesh generation again and filter out the final points. Experimental results indicate that the proposed simplification method is able to eliminate redundant data effectively. The reduction performance of this method is superior to the distance simplification method of Geomagic Studio obviously under the same required accuracy. It can be used in the real-time data acquisition process of 3D Reconstruction.