GUI Based Optical Character Recognition using Neural Networks

Abstract

The Optical Character Recognition (OCR) process means the transition from examined primer or written images to a machine- adjudged document. The American Standard Code for Information Interchange (ASCII) in cognitive processing uses OCR. The challenge is two primary clans word segmentation by letters and character recognition. Apply a new approach to include the two functions by Scale-Invariant Transforming Feature (SIFT) descriptors. To compare SIFT descriptors (RootSIFT), construct a new procedure, that offers outstanding results without accelerating calculation or repository conditions. In order to identify English characters, proposed system suggests that the reverse propagation neural network for bracket of character be employed. Conducted trials with further than 10 expedients aimed for every character and tried the delicacy for numerical figures, map letters, small letters, and alphanumeric symbols. The interpretation analysis of optimized neural network algorithm has attained an outside.