| 摘要 |
A method and a decoder for processing a digital video signal representing a sequence of images includes detecting motion pixels in a current image (I(t)), median filtering a transformed block (Bc) of the current image through a sub-step of spatial median filtering a first set of coefficients (Mcs, Mu, Md) and a sub-step of temporal median filtering a second set of coefficients (Mct, Mp, Mn), computing an inverse transformed block from the filtered block, and replacing pixels in the inverse transformed block by said detected motion pixels. |
