Kutatószeminárium

Thinning is a frequently used approach for skeletonization in a topology-preserving way. It is an iterative object reduction: some border points of binary objects are deleted until stability is reached. Another technique is distance-based skeletonization, which focuses on the detection of ridges or local maxima in the distance map of the input objects. In 3D, one of the three skeleton-like shape features is the medial surface - or so called surface skeleton. It should consist of 2D thin patches. Most 3D skeletonization algorithms work on digital pictures sampled on the cartesian cubic (CC) grid. Two alternative structures are the face-centered cubic (FCC) grid and the body-centered cubic (BCC) grid, which have beneficial topological and geometrical properties. According to our best knowledge, the latest surface skeletonization method working on the BCC or FCC grid was proposed in 2004. It combines distance-based approach and sequential thinning. We present two modified versions of this algorithm, which are faster and - according to our experiments - less sensitive to the visiting order of border points.