SVM based feature set analysis in dynamic malayalam handwritten character recognition

Abstract

Dynamic or Online handwritten character recognition is a challenging field in Human Computer Interfaces. The classification success rate of current techniques decreases when the dataset involves the similarity and complexity in stroke styles, number of strokes and stroke characteristics variations. Malayalam is a complex south Indian language spoken about 35 million people especially in Kerala and Lakshadweep islands. In this paper, a classification scheme based on support vector machines (SVM) is proposed to improve the accuracy in classification and recognition of online malayalam handwritten characters. SVM Classifier is a popular one in academy as well as in industry. This Classifiers are more suitable in a real world applicative problem, if we have major concern on the speed of recognition per character. The contribution of various features towards the accuracy in recognition is analyzed. Performance for different kernels of SVM are also studied. A graphical user interface has developed for reading and displaying the character. Different writing styles are taken for each of the 44 alphabets. Various features are extracted and used for classification after the preprocessing of input data samples. Feature Selection is carried out by choosing of different combinations of extracted features versus accuracy. Highest recognition accuracy of 97% is obtained for the best selected features in SVM with polynomial kernel. Recognition speed of a single stroke is obtained 0.52 secs.