Recognition of Printed Odia Characters and Digits using Optimized Self-Organizing Map Network

Optical character recognition is a kind of report picture investigation where filtered images that contain either machine printed or composed by hand content commitment to an optical character recognition (OCR) programming and making an interpretation of it into an editable machine discernible computerized content organization. An effort has been made to build up OCR framework for Odia language for that we have taken Self Organizing Map Network (SOM). Research in the field of character recognition for Odia content faces challenges generally in view of its properties like nature, various content styles and setting subordinate conditions of characters and their circumstance with respect to the check. This paper shows the impact of various element extraction strategy based on structural features are utilized in Odia character recognition framework with very high recognition accuracy rate. In this paper we proposed a Self Organizing Map Network (SOM) developed with specific structural features like height, width, cross section, end points to get high recognition accuracy up to 97.55%.