Recognition of Online Handwritten Gurmukhi strokes using support vector machine

Abstract

This paper presents an implementation to recognize Online Handwritten Gurmukhi strokes using Support Vector Machine. This implementation starts with a phase named Preprocessing, which consists of 5 basic algorithms. Prior to these algorithms, a basic step called Stroke Capturing is done, which samples data points along the trajectory of an input device. After preprocessing, recognition of Gurmukhi stroke is done using Support Vector Machine with the help of two cross validation techniques, namely, holdout and k-fold. The recognition is based on the unique IDs identified as the strokes in order to represent a Punjabi akshar (word). These strokes are taken from the one hundred Punjabi words written by 3 different writers. © 2013 Springer.