Non-rigid registration is a crucial step for many computer vision applications such as dimensional control in manufacturing, pattern recognition, etc. Generally, such application accuracy depends strongly on that of the registration step. In this paper, we propose a new method to determine an
efficient Coarse Registration that can be refined using Iterative Closest Point (ICP) algorithm. Firstly, two projection planes corresponding to the model and reference point clouds are determined using point dispersions. Secondly, all the points of each cloud are projected onto the corresponding planes. Then, two convex hulls are extracted from the two projected point sets
and then matched optimally. Next, the non-rigid transformation from the reference to the model is robustly estimated through minimizing the distance between the matched points pairs of the two convex hulls. Finally, this transformation estimation is refined using ICP algorithm. The experimental results based on synthetic data show that the proposed algorithm outperforms standard ICP in terms of execution time and precision whether with a normal or a large transformation.