3D-PatchMatch: An optimization algorithm for point cloud completion

Point cloud completion is an indispensible process for handling the occlusion problem occurring during the data acquisition. For completion, 3D point clouds generate a much larger searching space than 2D images while searching for the best match for boundary patches. To handle the searching speed bottleneck, this paper proposed a novel optimization algorithm which is called the 3D-PatchMatch algorithm for 3D point cloud completion. Inspired by the nature coherence of the point cloud data, the proposed algorithm use a random initialization process and a propagation process to reduce the convergence time. Furthermore, a random guess process is proposed to make sure the proposed algorithm has the ability of escaping from the local minima. Experimental results show significant time reduction rate for this method compared to brute-force search. Furthermore, the completion results with comparisons indicate that the proposed algorithm is rarely trapped in the local minima.