Shape decomposition for graph representation

The problem of shape analysis has played an important role in the area of image analysis, computer vision and pattern recognition. In this paper, we present a new method for shape decomposition. The proposed method is based on a refined morphological shape decomposition process.We provide two more analysis for morphological shape decomposition. The first step is scale invariant analysis. We use a scale hierarchy structure to find the invariant parts in all different scale level. The second step is noise deletion. We use graph energy analysis to delete the parts which have minor contribution to the average graph energy. Our methods can solve two problems for morphological decomposition - scale invariant and noise. The refined decomposed shape can then be used to construct a graph structure. We experiment our method on shape analysis.