3D Point Segmentation with critical point and fuzzy clustering

3D point or shape segmentation is an important technique and is widely used in computer graphics. In this work, we propose a new algorithm for 3D segmentation that is based on critical point identification and fuzzy clustering. There are two main goals for this proposal: First, to avoid over segmentation by finding the correct number of critical points automatically; second, divide object into meaningful components using fuzzy clustering. We apply our segmentation algorithm on some laser-scanned models to test its performance for automatic capture of geometric features in different data sets.