Partial Shape Retrieval by M-tree and a Bayesian Approach

An important problem in accessing and retrieving visual information is to provide efficient similarity machining in large databases. In this paper we propose partial retrieval by the shape similarity using Curvature Scale Space and an effective indexing using metric trees and a Bayesian approach. Each shape is partitioned into token attributes. Tokens are organized in a tree structure called M-tree. We focus on the study of the partial search ef ciency.