Certain investigations on improving the visual quality and psnr value using fuzzy logic in image restoration

Abstract

Image restoration is the process of improving the visual quality of the image by removing the unwanted information which is present in the image Image restoration plays a vital role in many fields such as computer vision medical field and research The unwanted information may include during capturing the image or transferring the image through Internet Here the unwanted information means changes in the brightness of the pixel and the value of the picture contrast This unwanted information is known as noise In many applications it is necessary to remove noise from image since the concert of other image processing tasks are reliant on the success of the image restoration process Noise removal is a tricky process in many image processing systems because it must not deform the needed information which is present in the image and saves the image details and texture while removing the noise Different types of noises are available They are speckle noise Gaussian noise Rayleigh noise Erlang noise exponential noise uniform noise and impulse noise Among these noises impulse noise is the most frequently occurred noise in images Impulse noise is also called as salt and pepper noise because it looks like a sprinkling salt and pepper These noises are removed by designing proper filters Filters are basically classified into linear filter and nonlinear filter Mean and median filters are well known nonlinear filters In mean filter each pixel is replaced by the mean value of neighboring pixel in the window size of 3 3 5 5 and 8 8etc. newline