Issue No. 02 - March/April (2009 vol. 24)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/MIS.2009.34
Paraskevi Tzouveli , National Technical University of Athens
Nikos Simou , National Technical University of Athens
Giorgios Stamou , National Technical University of Athens
Stefanos Kollias , National Technical University of Athens
The painters of Byzantine and post-Byzantine icons use specific rules and iconographic patterns. On the basis of these rules and patterns, a proposed knowledge-based image analysis system classifies icons. First, the system detects and analyzes the most important facial characteristics, providing rich, yet still imprecise, information about the icon. Then, it expresses the extracted information in terms of an expressive terminology formalized using a description logic (DL). To effectively handle the imprecision, the system uses a fuzzy extension of a DL to produce assertional knowledge. The combination of the terminological and assertional knowledge allows the system to categorize the icons. Tests of the system on a repository of 2,000 Byzantine icons has produced promising results.
knowledge representation formalisms and methods, artificial intelligence, scene analysis, image processing, computer vision, design methodology, pattern recognition, computing methodologies
P. Tzouveli, G. Stamou, N. Simou and S. Kollias, "Semantic Classification of Byzantine Icons," in IEEE Intelligent Systems, vol. 24, no. , pp. 35-43, 2009.