A Complete System for Bangla Handwritten Numeral Recognition

Abstract

This paper deals with a complete recognition system for unconstrained off-line Bangla (Bengali) handwritten numerals. To take care of variability involved in the writing style of different individuals, a robust scheme is presented here. The scheme is mainly based on new features obtained from the concept of water overflow from reservoir as well as topological and structural features of the numerals. The proposed scheme is tested on data collected from different individuals of various background and we obtained an overall recognition accuracy of about 92.8% from 12000 data.

Indexing terms:

• Handwritten segmentation
• Handwritten recognition
• Bangla numeral
• Indian script

Additional information

Notes on contributors

U Pal

U Pal received his PhD degree in Computer Science from Indian Statistical Institute in 1997. He has been with the Indian Statistical Institute, Kolkata from 1992 and is now a faculty member in the Computer Vision and Pattern Recognition Unit of the Institute. His fields of interest include Digital Document Processing, Optical Character Recognition (OCR), Medical Image Analysis, etc. In 1995, he received student best paper award from Computer Society of India. He also received a merit certificate from Indian Science Congress Association in 1996. He has about 75 research papers on various reputed journals and conference proceedings. Dr Pal has been serving as programme committee member of many national and international conferences. He is a life member of IUPRAI (Indian unit of IAPR) and Computer Society of India.

B B Chaudhuri

B B Chaudhuri received his PhD degree from Indian Institute of Technology, Kanpur in 1980. He joined Indian Statistical Institute in 1978 where he is the Head of Computer Vision and Pattern Recognition Unit of the Institute, currently on Sabbatical as a Jawaharlal Nehru Fellow. His research interests include Pattern Recognition, Image Processing, Computer Vision, Natural Language Processing, Speech Processing, Digital Document Processing and OCR. He pioneered the first Indian language Bharati Braille system as well as developed the first workable OCR for Bangla, Devnagari, Assamese and Oriya scripts. He has authored three books. He was awarded Sir JC Bose Memorial Award from IETE in 1986, MN Saha Memorial Award (twice) in 1989 and 1991, Homi Bhabha Fellowship award in 1992 for OCR and computer communication for the blind, Dr Vikram Sarabhai Research Award in 1995 for outstanding achievements in the fields of Electronics, Informatics and Telematics, C Achuta Menon Prize in 1996 for computer-based Indian Language Processing, Homi Bhabha Award: Applied Sciences from UGC in 2003 and Ram Lai Wadhwa Gold Medal from IETE in 2005. Also, he has been awarded the prestigious Jawaharlal Nehru Fellowship during 2004–2006. He is a member secretary (Indian Section) of International Academy of Sciences, Fellow of National Academy of Sciences (India), Fellow of Institution of Electronics and Telecommunication Engineering and Fellow of the Indian National Academy of Engineering. In 1998 he was elected as a Fellow of International Association of Pattern Recognition (IAPR). He has served as a guest editor of special issue of several journals including IETE Tech Review.

A Belaid

Abdel Belaïd received his PhD degree in Computer Science in 1979 and his DSc in 1987 from the University Henri Poincaré Nancy I, France. After a few years as Assistant Professor, he joined the National Centre for Scientific Research (CNRS) as a Research Scientist in 1984. In 2001, he became full Professor in University of Nancy 2 and responsible of the Cognitive Science Master. He leads since 1992 a research group at the LORIA working on Document Analysis and Text Recognition. His areas of research include Image Processing, Pattern Recognition, Document Analysis and Handwriting Recognition where he has authored over 150 articles which have been published in international journals and conferences. He developed handwriting recognition systems based on stochastic modeling, for linear and bi-dimensional representations.