Stroke based online handwritten Gurmukhi character recognition
Ramandeep Kaur, Mandeep Singh · 2016
In this paper, we present a preliminary system to effectively recognize the strokes for handwritten Gurmukhi characters. In this paper, 32 stroke classes have been considered and implemented for recognition. The proposed system extracts Spatiotemporal and Spectral features from collected stroke database. These features were then used to train the K-Nearest Neighbor (KNN), Multilayer Perceptron (MLP) and Support Vector Machines (SVM). The proposed methods for recognition were applied on the database using tenfold cross validation and percentage split technique. Recognition rate of 89.35% was obtained using K-Nearest Neighbor, 89.89% using Multilayer Perceptron and 89.64% using Support Vector Machines.