The current approaches in pattern recognition

The paper presents a brief survey of the current approaches in pattern recognition. In this field the classification processes of patterns are the main subject of interest. The most conventional techniques used to solve pattern recognition problems may be grouped into statistical, structural and hybrid methods of the previous two ones. The artificial intelligence approach to pattern recognition is used for such very complex tasks each solution of which heavily depends on the knowledge of experts. Recently, several attempts have been made to combine pattern recognition system with knowledge-based system so as to build a knowledge-based pattern recognition system of much more sophisticated recognition capability. The finks between the structural (syntactic) approach and the artificial intelligence one to pattern recognition are given.