Use of fuzzy feature descriptions to recognize handwritten alphanumeric characters

Abstract

Handwritten characters differ from person to person. Thus, when using traditional methods like neural networks and image processing techniques, extensive training of the system is needed. Due to this reason, an attempt was made to develop a system that used the methods that humans use to perceive handwritten characters. Thus, a system that recognizes features in handwritten characters using fuzzy logic was developed. The objectives in the development of this method were: 1. The development of a short and efficient algorithm that tries as much as possible to model human perception. 2. The development of an algorithm that can be implemented on any hardware or software platform through low computational power requirements. A time ordered sequence of coordinates should be sent into the system. The following calculations are then preformed on this set of data: 1. Segmentation of the image based on the direction of movement; 2. Determination of the universe of discourse; 3. Calculation of the local features for each identified segment; 4. Application to fuzzy rule-base for character recognition.