Similarity Judgment Method of Critical Point Graph Using Handwritten Character Recognition Technique

Abstract

We propose a similarity judgment method between critical point graphs (CPGs) in two volume data sets by using the directional element feature (DEF) that has been employed in the field of handwritten character recognition techniques (HCRT). A graph comparison technique called passing voxel method (PVM) which allows visualizing both similarity and differential of two CPGs. However, PVM has a drawback to the ability of its similarity judgment. On the other hand, CPG is composed of small straight-line segments. Therefore, we applied extended DEF to the similarity judgment of CPG, and solved the problem of PVM by using CPG with DEF.