Back to results
US 5,565,920 AGranted
Method and apparatus for video data compression using temporally adaptive motion interpolation
AssigneeThe Trustees of Princeton University
InventorsJungwoo Lee, Bradley W. Dickinson
Life of this patent
Estimating the term…
Abstract
In a system for compressing video data, the degree of global motion between a plurality of successive frames is determined for use in designating and spacing reference frames, relative to the global motion exceeding predetermined thresholds or levels of motion between certain ones of the frames.
Claims7
A system for compressing video data comprising video data associated with groups of pictures (GOP) including a predetermined number of frames, said system comprising:
- subband video coding means for receiving said video data, and extracting a plurality of spatial subbands from said video data, said spatial subbands as a whole representing subsampled pixels of individual frames, respectively;
- motion detection means for determining the degree of global motion between corresponding subsampled pixels of said spatial subbands of said frames, respectively;
- means responsive to said global motion measurements from said motion detection means for designating and adjusting the temporal spacing between reference frames;
- subband encoder means for encoding said spatial subbands of said reference frames, respectively;
- said designating means including means for coding certain of said reference frames as P frames relative to global motion between said spatial subbands;
- and
- said motion detection means including scene change detector means for coding said P frames with different bit numbers depending upon their relative motion content.
4 more elements
A method for compressing video data comprising the steps of:
- forming spatial subbands from said video data;
- determining the degree of global motion between frames of said video data by using selected ones of said spatial subbands of successive frames;
- adjusting the temporal spacing between reference frames relative to the degree of global motion measured between frames;
- coding certain of said reference frames as P frames with different bit numbers depending upon motion content, respectively, and
- selectively inserting said P frames into a frame sequence upon detecting a particular scene change.
2 more elements
Dependencies are inferred from the claim text.