Multi-level Image Classification Using Fuzzy Petri Net

Abstract

For a multi-level image classification, a knowledge representation scheme based on Fuzzy Petri Net with fuzzy inference algorithms is used. A simple graphical Petri net notation and a welldefined semantics displaying the process of reasoning through inference trees are used for visualization of the knowledge base and explanations of derived conclusion. Used knowledge representation formalism has the ability to show a probability of concepts and relations. The procedures of image multi-level classification using fuzzy recognition and inheritance algorithms on a knowledge representation scheme, as well as experimental results of image semantic interpretation, are presented.