A Review of Research on Optical Character Recognition for Devanagari language

Abstract

An Optical Character Recognition (OCR) System is a propelled framework by which embedded textual information is repossessed through digitizing from graphical media like an image. Developing such software system for handwritten characters with different handwriting is still a challenge. Sometimes it contains different letters in other than English language. This paper gives an entire idea about Optical Character Recognition (OCR) framework for computerized camera caught medias/pictures installed literary records. Toward the beginning, content regions are separated, and slant remedied. Around that level, the territories of one are binarized and divided into strains and characters. Noteworthiness module outperforms characters. Using a lot of one hundred endeavor card images to try various things, specialists have cultivated a most legendary accuracy of approx. 90%. As opposed to Tesseract, an open convey PC fundamentally based amazing OCR motor, present acknowledgment precision is really worth contributing. Besides, the unrivaled methodology is computationally unpracticed and expends low memory to be material available held gadgets.