3D reconstruction from multi-view point cloud based on particle system

Reconstructing accurate 3D object shapes from multi-view sampling points is one of the most challenging tasks to the scientists and engineers in reverse engineering. Aiming to this question, a merging and reconstructing algorithm based on particle system is presented for 3D topology reconstruction from multi-view point cloud in this paper. First one initial meshes of each point cloud in one view is formed using the Delaunay algorithm. Then an asymptotic performance is accomplished to deflate initial mesh towards the local concave boundary. Based on Newtonian law of motion and particle system, a physical constraint model of coupled particle systems is constructed, so new points are found and extracted and added into the initial mesh step by step, thus the initial mesh is shrank. At last several meshes are emerged into one mesh. In this paper, an experimental result is afforded and proves the usefulness of this method for reconstruction.