DCT image denoising: a simple and effective image denoising algorithm

This work presents an image denoising algorithm, arguably the simplest among all the counterparts, but surprisingly effective. The algorithm exploits the image pixel correlation in the spacial dimension as well as in the color dimension. The color channels of an image are first decorrelated with a 3point orthogonal transform. Each decorrelated channel is then denoised separately via local DCT (discrete cosine transform) thresholding: a channel is decomposed into sliding local patches, which are denoised by thresholding in the DCT domain, and then averaged and aggregated to reconstruct the channel. The denoised image is obtained from the denoised decorrelated channels by inverting the 3-point orthogonal transform.