AN EFFICIENT ALGORITHM FOR REMOVING SALT AND PEPPER NOISE FROM DOCUMENT IMAGES BY USING CONNECTIVITY PRINCIPLES

Abstract

The Proposing method improves the quality of degraded document images which are effected by salt and pepper noise. Noise effects the Document image quality and further processing. Thus to improve the quality of image we need to supress the noise. There are a number of methods to reduce different types of noise in document images. The proposing algorithm focusing on detecting and eliminating salt and pepper noise from infected document image. This proposing method works based on pixel connectivity
principles and labelling methods these methods helps to detect salt and pepper noise in the image. If pixels are connected in different label values then those pixels are connected and it is noise free. The results of the proposed technique are improved in terms of PSNR, SSIM, and MAE.