An Extension of "SEQ-NV-RANSAC" Approach To Avoid Bad-Segmentation Cases From Unstructured 3D Point Clouds Using Topology Information

Recently many applications require an automatic processing of massive unstructured 3D point clouds in order to extract planar surfaces of man-made objects. While segmentation is the essential step in feature extracting process, but bad-segmentation results (i.e. under and over-segmentation) are still standing as a big obstacle to extract planar surfaces with best fit reality. In this paper, we propose an extension of "SEQ-NVRANSAC" approach to avoid the bad-segmentation problems using topology information and intuitive threshold value. First, in order to avoid the under-segmentation problem, we check each one group which resulted from original "SEQ-NV-RANSAC" approach to get all neighbours points which have Euclidean distance less than the threshold value as a one surface group. This process will be repeated until no more points can be adding to that surface group. Then a new surface group will be created to check the remaining points. Second, in order to solve the oversegmentation, we propose three checks; the similarity of normal vectors (NV), the perpendicular distance and the intersection zone using bounding box test.