AN EFFICIENT APPROACH OF IMAGE RESTORATION USING MEAN INTENSITY CORRELATION OF NEIGHBORING PIXELS

Digital images are susceptible to the various types of noise. Noise in the image generally gets added due to the errors in the image acquisition process. These errors result in change in the intensity value of the pixel from the actual scene. Besides, this there are many other ways by which the noise can be introduced to an image. If an image is directly scanned from a photograph, then the grain noise is added into the image. Sometimes, noise gets added due to damage in the film. Motion of image capturing device introduces motion noise into an image. Moreover, electronic transmission of image data over a network can also introduce the noise to the image. There is various type of denoising methods available in the field of image restoration. However, most of these methods are computationally expensive and extensively complex to understand. Therefore, we have devised a very simple and efficient approach of image restoration using mean intensity correlation of neighboring pixel. Under this experiment, successful efforts have been made to restore a corrupted image with 85 percent reduction in the level of noise from the corrupt image.