Digital images are applied in various fields such us: physics, computer, engineering, chemistry, biology and medication sciences. It have been known that any images acquired by optical or electronic means is likely to be degraded by the sensing environment. Image restoration, is one of digital image processing field, which is care about improving the degraded image. Image restoration may be linear or non-linear and blind or non-blind. The following research focusing on linear non-blind image restoration and assuming that the degradation model as a convolution of the original image with blurring function and distorded by additive noise. Image restoration algorithms are trying to "undo" the blurring function and the noise from the degraded image by deconvolving the blurring function and reducing the noise from the degraded image to produce an estimate image, which it approach to the original image. The image have been used, blurred by Gaussian blurring function with selected standard deviation values = 1,2 and degraded by additive Gaussian noise with selected signal to noise ratio values SNR= 5, 10 and 20. The degradation have been used for three type of images, these are gray image (Satellite image), sonar image (Embryo image) and color image (bird image). Iterative Tikhonov-Miller filter and Wiener filter have been used to restore the degraded images. Using Root Mean Square Error (RMSE) measuring it have been concluded that, Iterative Tikhonov-Miller filter has better performance for less degradation parameters, with high SNR and Wiener filter has better performance for more degradation parameters, with low SNR.
