Pattern matching: overview, benchmark and comparison with F-transform general matching algorithm

The goal of this paper is to introduce a pattern matching problem and specify its role in the context of similar disciplines such as pattern recognition, content-based object retrieval and semantic object retrieval. Historical development of the disciplines together with a taxonomy is given and a short overview of currently used methods for pattern matching is presented. Moreover, a general soft computing algorithm for pattern matching based on fuzzy transform is proposed together with its version working over graphic cards. Finally, a comparison of the presented methods with respect to their possibilities of usage and computational time is given and graphically illustrated.