Feature Vector Table Based Image Binarization for Degraded Document Images

Abstract

Image is produced to record or display useful information but due to imperfections in the image capturing process and devices, the recorded image invariably represents a degraded version of the original. These degradations may be as a result of addition of noise, geometrical degradations, illumination and blur. These problems degrade the final quality of the acquired images for display and thus complicate the task of finding the error in object recognition and event detection. The primary objective of this paper is to develop a method for filtering, clearing the historical and degraded documents. ‘Wiener filter algorithm' use to clear degraded historical handwritten documents .Experimental results is calculated by using MSE(i.e. mean square error) and PSNR (i.e. peak signal noise ratio).This algorithm helps to get the maximum accuracy and performance.