Abstract:

       Image restoration is a key issue in digital image processing. Image quality will decline due to camera defous, relative motion between a camera and an objective and noise. This paper presents partition based image restoration approach with improved Wiener filtering for the constraint of classical Wiener filtering. The method can not only perform pseudo morph attenuation and noise suppression, but also preserve the edge of an image. We evaluate the signal to noise ratio (SNR) of the peak values of a processed image by the two methods. Experimental results show that the peak SNR of an image processed by improved restoration method is higher than that of an image processed by the classical method.

Key words: image restoration, motion blur, defocus blur, Wiener filtering