Online Recognition of Malayalam Scripts with Minimized Feature Dimensions

Abstract: The objective of the work is to develop a database for online Malayalam script and identify online handwritten Malayalam characters with the use of Artificial Neural Network (ANN) and K-Nearest Neighbour (KNN) classifiers. The system uses minimum features (ADP, Aspect Ratio, Intersections and Octants)identified as relevant features and obtained 90.40% accuracy which is observed to be the best performance with these features and classifiers. Experiments are done for handwritten basic vowel alphabets (8) and consonants (36) in Malayalam. The database consists of 12800 samples of 64 Malayalam characters with 200 samples per character.

Keywords: Script Database; OHCR; KNN; ANN.