3-D mean-separation-type short-time DFT with its application to moving-image denoising

Although for a still image the 2-D DFT and the 2-D DCT have similar properties to each other, for a moving-image sequence the 3-D DFT gets an advantage of representing the sequence more compactly over the 3-D DCT. Through the mathematical analysis of the 3-D DFT and the 3-D DCT based on a simple signal model of a moving-image sequence, this paper shows that the even symmetrization employed implicitly by the 3-D DCT may cause deterioration of representation efficiency and hence the 3-D DFT can achieve better representation efficiency than the 3-D DCT. In addition, to improve the suitability of the 3-D short-time DFT to processing of video signals which generally have significant local DC components carrying important structural information, this paper introduces a technique of local-mean-separation as a preprocess of the 3-D short-time DFT, thus to construct 3-D mean-separation-type ST-DFT; applies it to video denoising, and demonstrates its advantage over the existing 3-D transforms through experimental simulations of video denoising.