Recognition of Handwritten Tamil Characters Using Statistical Classifiers

This paper describes a system to recognize handwritten Tamil characters using statistical classifier approach, for a subset of the Tamil alphabet. The process of handwriting recognition involves extraction of some defined characteristics called features to classify an unknown handwritten character into one of the known classes. A typical handwriting recognition system consists of several steps, namely: preprocessing, segmentation, feature extraction and recognition. Data input (the hand written Tamil character) were collected from different writers on A4 sized documents. They were scanned using a flat-bed scanner at a resolution of 100 dpi and stored as 256-bit color scale images. We trained the system with 500 characters belonging to 10 characters. The testing data contained a separate set of 50 characters. A training data was used to test the system, to see how well the system represents the data it has been trained on. In the test set, a recognition rate of 90% was achieved.