Recognition of Online Handwritten Devanagari Numerals using Support Vector Machine

Abstract

Handwriting has continued to persist as a mean of communication and recording information in day-to-day life even with the invention of new technologies. Natural handwriting is one of the easiest ways of information exchange between a human and a computer. Handwriting recognition has attracted many researchers across the world since many years. Recognition of online handwritten Devanagari numerals is a goal of many research efforts in the pattern recognition field. The main goal of the work presented in this dissertation is the recognition of online handwritten Devanagari numerals using support vector machine. In the data collection phase, co-ordinate points of the input handwritten numeral are collected as the numeral is written; various algorithms for pre-processing are applied for normalizing, resampling and interpolating missing points, smoothing and slant correction. Two low-level features i.e. direction angle and curvature are extracted from the pre processed data. These features along with the x and y coordinates of the input handwritten character are stored in a .csv file and fed directly to the recognition phase. Recognition is done using four kernel functions of SVM by partitioning the data into different schemes. The recognition accuracies are obtained on different schemes of data using the four kernel functions of SVM for each scheme.