Universiti Teknologi Malaysia Institutional Repository

Feature extraction in holistic approach for arabic handwriting recognition system: a preliminary study

Al-nuzaili, Qais and Mohamad, Dzulkifli and Ismail, Nor Azman and Khalil, Mohammed S. (2012) Feature extraction in holistic approach for arabic handwriting recognition system: a preliminary study. In: Proceedings - 2012 IEEE 8th International Colloquium on Signal Processing and Its Applications, CSPA 2012. IEEE, New York, USA, pp. 335-340. ISBN 978-146730961-5

Full text not available from this repository.

Official URL: http://ieeexplore.ieee.org/document/6194745/

Abstract

Handwriting recognition is the ability of a computer to receive and interpret intelligible handwritten input. Recognition systems are divided into two categories: holistic approach and analytical approach. A holistic approach handles the whole input image, while analytical approach involves two steps namely; segmentation and combination. Handwriting recognition began long time ago mainly in Latin and Chinese characters. However, little effort has been devoted to Arabic characters. The domain of handwriting in the Arabic script presents unique technical challenges and has been given more attention recently than other domains. In respect to the above issue, this paper investigates two different feature extraction methods, Angular span method and Distance span method, which may represent the distribution of pixels in the word properly. Samples from IFN/ENIT benchmark dataset are used to evaluate both methods.

Item Type:Book Section
Additional Information:Indexed by Scopus
Uncontrolled Keywords:arabic handwriting recognition, feature extraction, holistic approach
Subjects:P Language and Literature > P Philology. Linguistics
Divisions:Computer Science and Information System (Formerly known)
ID Code:35777
Deposited By:INVALID USER
Deposited On:11 Nov 2013 17:52
Last Modified:13 Jun 2017 11:46

Repository Staff Only: item control page