A complete OCR system for continuous Bengali characters

Author

Mahmud, J.U. ; Raihan, Mohammed Feroz ; Rahman, Chowdhury Mofizur

Author_Institution

Dept. of Comput. Sci. & Eng., Bangladesh Univ. of Eng. & Technol., Dhaka, Bangladesh

Volume

4

fYear

2003

fDate

15-17 Oct. 2003

Firstpage

1372

Abstract

This paper is concerned with a complete optical character recognition (OCR) system for Bengali character. Recognition is done for both isolated and continuous printed multi font Bengali characters. Preprocessing steps includes segmentation in various levels, noise removal and scaling. Freeman chain code has been calculated from scaled character which is further processed to obtain a discriminating set of feature vectors for the recognizer. The unknown samples are classified using feed forward neural network based recognition scheme. It has been found from experimental results that success rate is approximately 98% for isolated characters and 96% for continuous character.

Keywords

backpropagation; feature extraction; feedforward neural nets; image classification; image segmentation; optical character recognition; Freeman chain code; OCR system; back propagation; continuous Bengali characters; feature extraction; feature vectors; feed forward neural network; noise removal; optical character recognition; scaling; segmentation; Character recognition; Computer science; Feature extraction; Image recognition; Isolation technology; Neural networks; Noise level; Optical character recognition software; Optical noise; Paper technology;

fLanguage

English

Publisher

ieee

Conference_Titel

TENCON 2003. Conference on Convergent Technologies for the Asia-Pacific Region

Print_ISBN

0-7803-8162-9

Type

conf

DOI

10.1109/TENCON.2003.1273141

Filename

1273141