Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Variable-raster multiresolution video processing with motion compensation techniques

: Ohm, J.-R.; Rummler, K.


IEEE Signal Processing Society:
International Conference on Image Processing. Proceedings. Vol.1
Piscataway, NJ: IEEE, 1997
ISBN: 0-8186-8183-7
ISBN: 0-8186-8184-5
ISBN: 0-8186-8185-3
International Conference on Image Processing (ICIP) <1997, Santa Barbara/Calif.>
Fraunhofer HHI ()
band-pass filters; image reconstruction; image resolution; image sampling; motion compensation; video signal processing; wavelet transformation; variable-raster multiresolution video processing; standard format; video acquisition; video transmission; video presentation; interlaced scanning; digital video services; progressive format; signal resolution; temporal sampling effects; vertical sampling effects; motion-compensated vertical filters; sampling rate conversion; progressive raster; interlaced raster; polyphase filter; factor-2 upsampling filterbank; factor-2 downsampling filterbank; polyphase shift; motion parameters; multiresolution representation; video data; spatio-temporal subband framework; wavelet framework; perfect reconstruction filters

The standard format for video acquisition, transmission and presentation uses an interlaced scanning technique. On the other hand for digital video services, the progressive format is more desirable in many situations, and different resolutions of the signal are required. The interlaced scan is based on a tight concatenation of vertical and temporal sampling effects. This paper introduces the technique of motion-compensated vertical filters within pairs of fields to solve the problem of sampling rate conversion between interlaced and progressive raster. It is shown that the motion-compensated vertical filter is a polyphase realization of a factor-2 up- or downsampling filterbank, with polyphase shift controlled by motion parameters. Possible applications are down- and up-conversion between progressive and interlaced formats, and multiresolution representation of video data within a spatio-temporal subband or wavelet framework.