Open Access
ARTICLE
Devanagari Handwriting Grading System Based on Curvature Features
Department of Computer Applications, GZS Campus College of Engineering & Technology, (Maharaja Ranjit Singh Punjab Technical University), Bathinda, Punjab, India. E-mail: munishcse@gmail.com
Computer Science and Engineering Section, Yadavindera College of Engineering, Talwandi Sabo, Bathinda, Punjab, India. E-mail: simpel_jindal@rediffmail.com
Computer Modeling in Engineering & Sciences 2017, 113(2), 195-202. https://doi.org/10.3970/cmes.2017.113.201
Abstract
Grading of writers in perspective of their handwriting is a challenging task owing to various writing styles of different individuals. This paper presents a framework for grading of Devanagari writers in perspective of their handwriting. This framework of grading can be useful in conducting the handwriting competitions and then deciding the winners on the basis of an automated process. Selecting the set of features is a challenging task for implementing a handwriting grading system of particular language. In this paper, curvature features, namely, parabola curve fitting and power curve fitting have been considered for extracting the vital information of writers, which can be helpful in grading the writers. For obtaining the classification score, k-NN classifier has been considered in the present work. Four printed Devanagari font styles, namely, Devlys, Krishna, Krutidev, and Utsaah have been considered for train the proposed model of handwriting evaluation. For evaluating the effectiveness of the proposed framework, we have conducted a mock test of 75 Devanagari writers (Left handed and Right handed) and noticed that the proposed framework performing well suitable for conducting the handwriting competition of Devanagari text writers. This work is also newly in the ground of Devanagari text recognition.Keywords
Cite This Article
This work is licensed under a Creative Commons Attribution 4.0 International License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.