CIDER: Corrected Inverse-Denoising Filter for Image Restoration

摘要

In this paper we propose and develop a new algorithm, Corrected Inverse-Denoising filtER (CIDER) to restore blurred and noisy images. The approach is motivated by a recent algorithm ForWaRD, which uses a regularized inverse filter followed by a wavelet denoising scheme. In ForWaRD, the restored image obtained by the regularized inverse filter is a biased estimate of the original image. In CIDER, the correction term is added to this restored image such that the resulting one is an unbiased estimator. Similarly, the wavelet denoising scheme can be applied to suppress the residual noise. Experimental results show that the performance of CIDER is better than other existing methods in our comparison study.