A Novel Blind Restoration Technique Using Different Edge Detection Methods

Siddharth Saxena; Rajeev Kumar Singh

doi:10.37591/joosdt.v1i2.132

Abstract

Image restoration is a field of image processing in which we recover an original image from a degraded image. Objective of this paper is to compare different blurred and noisy images which are reconstructed with the help of blind deconvolution algorithm using different edge detection methods such as Canny, Sobel, Prewitt, Roberts, Laplacian. Initially original image is blurred using Gaussian filter. After that in the edges of blurred image ringing effect can be detected using different edge detection methods then it can be removed before restoration process. After that blind deconvolution algorithm is applied to get the reconstructed image. We used peak signal to noise ratio (PSNR) value for the comparison of different deblurred images. The higher the PSNR value, the better the image. Generally it is believed that blind deconvolution algorithm gives better result when there is no information about the distortion. It restores the image and point-speard function (PSF) at the same time.Keywords: Blind deconvolution algorithm, edge detection methods, image restoration, PSF