A Tabulation method for Character Recognition using Haar wavelet in 2D images

Abstract

Character perception is a unique way to diagnostic the unidentifiable character into understandable text. Nowadays, with the increasing urge of getting information digital media has been roped in because of its convenient use. Hence the printed media needs to get into digital format but the printed text needs to get recognized by the computer this difficulty is solved by the pattern matching technique and haar wavelet method to recognize the character ,decompose it and frame it into understandable format keeping its prototype in mind this method can turn boon for the unidentifiable text also it can overcome the problems of printed text character perception system help human ease and reduce their jobs of manually handling and processing of document. The proposed method showed that the recognition of text is reliable and significantly accelerates data processing. Evaluation is executed as a collection of degraded documents which are images, indicates the proposed work will be well suited to reduce the noise by improving the pixel quality in which it will be rendered. The result analysis performed by using image quality matrix which are Mean Squared Error, Signal to Noise ratio, Structural Content, Normalized Cross Correlation, Maximum Difference, Normalize Absolute Error, Average difference. All the parameter shows effective results.