A Novel Algorithm for Analysis of a Local Shape In the 3-D Gray Image

TLDR: This paper discusses about topology preserving 3-D skeletonization sequential algorithm which computes curve skeletons for the given solid object, which does not require any computations which are complex, large number of windows which make implementation a complex.

Abstract: For shape analysis in 3-D, probably skeletonization will become a momentous tool for analyzing the different shapes in 3-D, as similar to the 2-D. This paper discusses about topology preserving 3-D skeletonization sequential algorithm which computes curve skeletons for the given solid object. The original geometry of the given object will be well-preserved by curve skeleton. In the proposed algorithm, each voxel of the three-dimensional image is classified as two classes: boundary, corner. The object is scanned with the use of a structuring element, border and corner voxels are classified based on the certain condition proposed in this paper. This algorithm does not require any computations which are complex, large number of windows which make implementation a complex. The proposed algorithm is various 3-D simulated images and resulting skeletons are found to satisfactory. The 3-D simulated images and its skeletons are viewed with the help of ImageJ software.