A novel intelligent system for defining similar symbols

We introduce an expert system for the recognition of any typed or handwritten symbols, then, we describe how a symbol can be represented by another symbol which is formed of only straight line segments. This allows a large number of different styles of handwritten or typed symbols to be mapped into a much smaller number of representations. These representations are used as models for the automatic recognition of symbols. The system uses the structural pattern recognition technique for modeling symbols by a set of straight lines referred to as line segments. The system rotates, scales and thins the symbol, then extracts the symbol strokes. Each stroke is then mapped into line segments. The system is shown to be able to map similar styles of the symbol to the same representation. When the system had some stored models for each symbol (an average of 97 models/symbol), the recognition rate was 95%, and the rejection rate was 16.1%. The system was tested by 5726 handwritten English characters from the Center of Excellence for Document Analysis and Recognition (CEDAR) database.