A simple and effective technique for online handwritten bangla character recognition

Abstract

In this paper, we have proposed a simple but effective feature extraction technique following the distance-based features to recognize online handwritten isolated Bangla basic characters. In this approach, a character is divided into N number of segments and then distances are calculated among each other. These distance values are then used as features for recognition purpose. On evaluation of this feature set on 10,000 Bangla character samples (50-class character set) by various classifiers, the method yields reasonably good result with 98.20% success rate.