Restoration of randomly blurred images by the Wiener filter

The restoration of images distorted by systems with noisy point spread functions and additive detective noise is considered. The criterion considered for the restoration is based on the Wiener technique. The proposed Wiener-based filter is iterative in nature. The overall computation of this modified Wiener filter can be carried out in the frequency domain using the fast Fourier transform and circulant matrix approximation. Experimental results show that the modified Wiener filter outperforms its linear counterpart (neglecting the impulse-response noise). The modified Wiener filter also gives better restoration results than the Backus-Gilbert technique. The Wiener-based filter is found to be computationally robust and inexpensive.<
>