Extreme video compression over a fixed bandwidth channel

1. A method of encoding a video, comprising:

first, carrying out first motion compensation within a video that has resulted from camera motion, by first finding a group of frames, and finding camera motion among the group of frames, and carrying out first motion compensation within said group of frames that has resulted from camera motion to shift frames in said group of frames in a way that creates a smallest mean square error, and to reduce said error to align the group of frames into a stack, wherein said motion compensation finds borders of said frames as part of said first motion compensation, and wherein said borders are not motion compensated;

second, carrying out second motion compensation in the stack that is not resulting from camera motion; encoding said first and second motion compensations separately; and

dividing the video into first and second parts, which respectively occupy first and second parts of a total bit budget for compression, separately compressing the first part to create a first compressed part as said first carrying out, and to create a second part to create a second compressed part as said second carrying out, wherein said first compressing and said second compressing use different compression techniques.

2. A method as in claim 1 , wherein said encoding comprises separately using spatial compression and temporal compression.

3. A method as in claim 1 , wherein said first part uses said camera motion compensation and said second part uses a general motion compensation.

4. A method as in claim 3 , wherein said first part uses 75% of an available bit budget and said second part uses 25% of the available bit budget.

5. A method as in claim 1 , wherein said first encoding uses a downsampling technique which downsamples a video by an amount η, where η2 is a numerical percentage of downsampling.

6. A method as in claim 5 , wherein said first encoding downsamples the video by four, by removing every second line.

7. A method as in claim 1 , further comprising decomposing the video into a group of frames.

8. A method as in claim 7 , wherein said decomposing comprises comparing each frame in the group of frames with another frame in the group of frames.

9. A method as in claim 8 , wherein said comparing comprises comparing each frame to a frame in the forward direction and also comparing each frame to a frame in the backward direction.

10. A method as in claim 1 , wherein said encoding includes all of temporal compression, spatial compression and entropy compression.

11. A method as in claim 1 , wherein said encoding comprises decomposes an image into different resolutions arranged hierarchically from coarsest detail of resolution to finest detail of resolution.

12. A encoding system, comprising:

a computer, carrying out compression of a video, by first finding a group of frames, and finding camera motion among the group of frames, and carrying out first motion compensation within said group of frames that has resulted from camera motion to shift frames in said group of frames in a way that creates a smallest mean square error, and to reduce said error to align the group of frames and also assembling the frames which have been aligned into a stack, and separately carrying out second motion compensation in the video that is not resulting from camera motion and separately encoding and compressing data resulting from said first motion compensation and said second motion compensation;

wherein said computer finds borders of said frames as part of said first motion compensation, and wherein said borders are not motion compensated; and

divides the video into first and second parts, which respectively occupy first and second parts of a total bit budget for compression, and separately compresses the first part to create a first compressed part, and to create a second part to create a second compressed part, wherein said first compressing and said second compressing use different compression techniques.

13. A system as in claim 12 , wherein said encoding system separately uses both spatial compression and temporal compression.

14. A system as in claim 12 , wherein said encoding system uses a downsampling technique which downsamples a video by an amount η, where η 2 is a numerical percentage of downsampling.

15. A system as in claim 14 , wherein said first encoding downsamples the video by four, by removing every second line.

16. A system as in claim 12 , wherein said encoding system decomposes the video into a group of frames.

17. A system as in claim 16 , wherein said decomposes comprises comparing each frame in the group of frames with another frame in the group of frames.

18. A system as in claim 17 , wherein said comparing comprises comparing each frame to a frame in the forward direction and also comparing each frame to a frame in the backward direction.

19. A system as in claim 12 , wherein said encoding system uses all of temporal compression, spatial compression and entropy compression.

20. A system as in claim 12 , wherein said encoding system uses a wavelet transform that decomposes an image into different resolutions arranged hierarchically from coarsest detail of resolution to finest detail of resolution.

21. A method of encoding a video, comprising: encoding a video to create a graded resolution portion, having a first image portion representing a coarse level of details, and a second resolution portion representing a second finer level of details;

spatially encoding a portion of the video and also temporally encoding said video, wherein said temporally encoding creates a first encoded portion that first motion compensation within a video that has resulted from camera motion among a group of frames, a border location of said group of frames of a video, and has second motion compensation in the video that has not resulted from camera motion, and wherein said second motion compensation does not encode said border location, and also entropy encoding said video; and

dividing the video into first and second parts, which respectively occupy first and second parts of a total bit budget for compression, separately compressing the first part to create a first compressed part as said first carrying out, and to create a second part to create a second compressed part as said second carrying out, wherein said first compressing and said second compressing use different compression techniques.

22. A method as in claim 21 , further comprising downsampling said video by an amount η, where η 2 is a numerical percentage of downsampling.

23. A method as in claim 22 , wherein said first encoding downsamples the video by four, by removing every second line.