Evaluating Tree-Shaped Network Similarity via Morphological Feature Matching With Applications

Abstract

The structure of the tree is presented in many applications both in natural and man-made forms. In reality, various structures of the human body have a branching configuration similar to the morphology of tree. The retinal vascular tree, brain stem, and nervous system are examples of tree-shaped systems. Similarity analysis of such tree-shaped networks is a vital task in many applications that require shape matching(s) with minimum effort and time. Thus, this paper proposes an approach for evaluating the degree of tree-shaped network similarity through morphological feature matching. This study focuses on the graphical representation of tree-shaped morphological parameters. The degree of similarity is determined by considering one morphological parameter based on the number of branches per order. The results of testing the proposed technique on a selected set of tree-shaped networks have shown the great efficacy of the approach in yielding the degree of similarity. Finally, the proposed method has unlimited scope of applications handling the structure of tree such as water basins, electric power generation, etc. for the aggregation system, and roads, electric power distribution networks, etc. for the disaggregation system.