Denoising by averaging reconstructed images: Using Singularity Function Analysis

A newfound method of denoising that based on Averaging Reconstructed Image (AVREC), is used. The approach was proposed on signals, approximately about last decade, Since 2004. In definition (procedure), first of all, we divide the spectrum of noisy image into several images that can be then, reconstructed with 2-D Singularity Function Analysis (SFA) model. Among this mathematical model, each matrix or a discrete set of data, represents as a weighted sum of singularity functions. In image denoising field, this technique, rebuilt all lost high frequencies parameters that are essential. Illustrate each new image, as a sum of noise-free image and the small noise. So on, we can then, denoise image by averaging reconstructed ones. Both theoretical and experimental results on standard gray-scale images, confirm the advantages (benefits) of this approach as an applicable method of denoising.