10-10-2014, 09:23 AM
Abstracts: In the field of Image processing, OCR is one of the leading and developing idea. There are myriad reasons why this idea made it compelling to work on. Due to human limitation, variety of languages world-wide, handwriting difference from person to person, font styles etc. this field is now developing day by day. Researches and implementation of English OCR has already been done time ago. India being a multi lingual country needed such a medium that could identify and recognize characters of many different languages used in day to day life. Hindi being the national language is the most used across India. Government documents, Banks, Literatures, Books, Court documents etc. all make use of Hindi script. Also in a country with such huge population makes document managing and preservation difficult. Hence, this project presents an efficient algorithm for recognition of Hindi script characters from printed and handwritten documents. In this study, referential approach has been implemented on some printed documents and results have been obtained successfully. The COR system is improved by incorporating a couple of pre-processing steps, segmentation, recognition and finally post processing.