AN OVERVIEW OF HANDWRITTEN GURMUKHI CHARACTER RECOGNITION

Abstract- In this thesis work we have suggested offline recognition of isolated handwritten characters of Gurmukhi script. We have also prolonged the work by applying the same methodology to recognize handwritten Gurmukhi numerals. In our work we have considered 35 fundamental characters of Gurmukhi script all assumed to be isolated and bearing header lines on top to recognize. In numerals, handwritten elements of ten digits from different writers are considered. Gurmukhi numerals using three characteristic sets and three classifiers. Among three characteristic sets, first characteristic set is comprised of distance profiles having 128 characteristic. Second characteristic set is comprised of different types of projection histograms having 190 characteristic. Third characteristic set is comprised of zonal density and Background Directional Distribution forming 144 features. The three classifiers used are:- SVM, PNN & K-NN The SVM classifier is used with Radial Basis Function kernel. We have observed the 5-fold cross verification accuracy in the case of each characteristic set and classifier. We have obtained the optimized result with each combination of characteristic set and classifier by adjusting the different limits. The results are compared and trends of result in each combination of characteristic set and classifier with varying limits is also discussed. With PNN and K-NN the highest results are obtained using third characteristic set as 98.33 % and 98.51 % respectively while with SVM the highest result is obtained using second characteristic set as 99.2%. The results with SVM for all characteristic sets are higher than th