Point sets simplification using local surface analysis

This paper studies the problem of simplifying densely distributed point-sampled models. Many computer graphics applications call for vivid, full detail models. However, the level of detail necessary is more important than this need for fidelity in rendering system. So it is useful to obtain simple versions of complex models. We have developed a novel simplification algorithm which can preserve the shape information such as sharp corner of the original point set. The underlying simplification principle is based on the local surface analysis of each sample point. Our method decimates only one point every time from the point-sampled models to be simplified. The algorithm supports the generation of progressive and multiresolution expression of the input point set, and thus can be applied to progressive transmission over a network with limited bandwidth. The effectiveness and performance of the proposed method are validated and illustrated through case studies.