Temporal filtering of wavelet-compressed motion imagery

Temporal filtering of motion imagery can alleviate the effects of noise and artifacts in the data by incorporating observations of the imagery data from several distinct frames. If the noise that is expected to occur in the data is well-modeled by independent and identically distributed (IID) Gaussian noise, then straightforward algorithms can be designed to filter along motion trajectories in an optimal fashion. This paper addresses the restoration of motion imagery that has been compressed by scalar quantization of the data´s two- or three-dimensional discrete wavelet transform coefficients. Noise due to compression in such situations is neither independent nor identically distributed, and thus simple filters designed for the IID case are suboptimal. This paper shows how proper statistical characterization of the compression error can be used in temporal filtering to improve the visual quality of the compressed motion imagery.