Investigation of normalization techniques and their impact on a recognition rate in handwritten numeral recognition

Investigation of normalization techniques and their ...

title:	Investigation of normalization techniques and their impact on a recognition rate in handwritten numeral recognition
author:	Chmielnicki Wiesław , Stąpor Katarzyna
Journal title:	Schedae Informaticae
volume:	19
date of publication:	2010
pages:	53-77
ISSN:	1732-3916
eISSN:	2083-8476
URL:	http://www.ejournals.eu/sj/index.php/SI/article/view/3000
date accessed:	2019-06-10
language:	English
Journal language:	English
abstract in english:	This paper presents several normalization techniques used in handwritten numeral recognition and their impact on recognition rates. Experiments with five different feature vectors based on geometric invariants, Zernike moments and gradient features are conducted. The recognition rates obtained using combination of these methods with gradient features and the SVM-rbf classifier are comparable to the best state-of-art techniques.
keywords in english:	handwritten numeral recognition, normalization techniques, SVM classifier, feature vectors, OCR, geometric invariants, Zernike moments, gradient features
affiliation:	Wydział Fizyki, Astronomii i Informatyki Stosowanej : Zakład Technologii Gier
type:	journal article
subtype:	academic paper