Method and apparatus for selectively encoding and decoding a digital motion video signal at multiple resolution levels

• US 5,122,873 A
• Filed: 02/19/1991
• Issued: 06/16/1992
• Est. Priority Date: 10/05/1987
• Status: Expired due to Term

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Golin, Stuart J.
• Primary Examiner(s): Kostak, Victor R.
• Application Number:
  US07/657,708
• Time in Patent Office: 483 Days
• Field of Search: 358/133, 358/138, 358/136, 358/13, 382/49
• US Class Current: 375/240.23
• CPC Class Codes: H04N 19/00 Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
• CPC Class Codes: H04N 19/10 using adaptive coding
• CPC Class Codes: H04N 19/186 the unit being a colour or a chrominance component
• CPC Class Codes: H04N 19/50 using predictive coding H04N19/61 takes precedence
• CPC Class Codes: H04N 19/53 Multi-resolution motion estimation; Hierarchical motion estimation
• CPC Class Codes: H04N 19/61 in combination with predictive coding
• CPC Class Codes: H04N 19/63 using sub-band based transform, e.g. wavelets
• CPC Class Codes: H04N 5/85 on discs or drums
• CPC Class Codes: H04N 5/9262 using predictive coding
• CPC Class Codes: H04N 7/24 Systems for the transmission of television signals using pulse code modulation H04N21/00 takes precedence
• CPC Class Codes: H04N 9/8066 with insertion of the PCM audio signals in the vertical blanking interval of the PCM video signal

Abstract

At least one selected image of a digital video signal is encoded at multiple levels of resolution. For each level of resolution, a correction image is formed by subtracting the value of each pixel in a reference image of that resolution level from the value of each corresponding pixel in the selected image of that resolution level. The correction image is quantized and encoded. A decoder decodes the encoded quantized correction image for each resolution level. The value of each pixel in the decoded correction image having the lowest resolution is added to the value of each corresponding pixel of a reference image having the same resolution to form a result image of the lowest resolution. This result image is expanded to the next higher level of resolution. The value of each pixel in the expanded result image is added to the value of each corresponding pixel in the decoded correction image of the same resolution level to form a result image of that resolution level. This process is continued until each pixel value in the expanded result image, which has been expanded to full resolution, is added to the value of each corresponding pixel in the full resolution decoded correction image to form the final image which is stored in memory for subsequence display.

Claims

1. A method of encoding a digital motion video signal comprising the steps of:

a--selecting at least one frame from a sequence of frames of said digital motion video signal;

b--resolving each selected frame into at least one lower level of resolution;

c--encoding at least one lower resolution level of each of said selected frames; and

d--encoding each non-selected frame of said digital motion video signal using only the full resolution level.

2. The method in accordance with claim 1 wherein step c comprises encoding at least one lower resolution level but not encoding the full resolution level of each of said selected frames.

3. The method in accordance with claim 1 in which step c comprises, for each level of resolution of said selected frame which is encoded, the steps of:

i--forming a correction image comprising an array of pixel values, by subtracting pixel values in a reference image from corresponding pixel values in that level of resolution of said selected frame;

ii--quantizing the pixel values in said correction image to form a quantized correction image; and

iii--encoding said quantized correction image.

4. The method in accordance with claim 3 in which step i for the lowest level of resolution of said selected frame includes the step of providing a reference frame comprising an array of pixels all of which have the same value.

5. The method in accordance with claim 4 in which step i for each level of resolution higher than the lowest level of resolution of said selected frame includes the steps of:

(1) forming a result image having a lower level of resolution by adding the pixel values of the quantized correction image at that lower level of resolution to the pixel values of the reference image at that lower level of resolution; and

(2) forming the reference image at the higher level of resolution by expanding the result image formed in step (1) to said higher level of resolution.

6. The method in accordance with claim 3 wherein step iii comprises the steps of dividing said quantized correction image into regions and generating region parameters descriptive of the position, size and fill data for each region.

7. The method in accordance with claim 6 wherein the step of dividing said quantized correction image into regions comprises dividing said image into null and non-null regions or into null and linear fill regions.

8. The method in accordance with claim 7 wherein said region descriptions are encoded using binary tree decomposition.

9. The method in accordance with claim 8 in which said non-null regions are encoded using vector quantization or discrete cosine transform techniques.

10. The method in accordance with claim 9 wherein encoding using vector quantization comprises encoding using DYADS, QUADS, or DPCM.

11. The method in accordance with claim 1 in which step d comprises the steps of:

i--dividing each non-selected frame into regions and generating region parameters descriptive of the location and size of respective regions within said frame;

ii--for non-selected still frames or a first frame of a sequence of non-selected frames;

(1) determining absolute fill data describing respective regions;

(2) determining DPCM fill data describing respective regions for which absolute fill data cannot be determined;

(3) variable length encoding said absolute fill data, said DPCM fill data and corresponding region descriptive parameters;

(4) determining an estimate of decompression time from said variable length encoded absolute, DPCM fill data and corresponding region descriptive parameters;

(5) providing said variable length encoded fill data and corresponding region parameters if said estimate is less than a predetermined value, and if not, repeating steps (i) and (1)-(5) using a different tolerance in determining said absolute fill data; and

iii--for a succession of non-selected frames;

(1) comparing pixel values in each region with corresponding pixel values in a like sized region from a previously occurring frame to develop respective pixel differences;

(2) determining relative fill data descriptive of pixel differences for respective regions;

(3) determining DYAD fill data for respective regions in which relative fill data cannot be determined;

(4) determining DPCM fill data for respective regions in which relative fill data cannot be determined and for which DYAD fill data cannot be determined; and

(5) variable length coding said relative fill data, said DYAD fill data, said DPCM fill data and corresponding region parameters.

12. The method in accordance with claim 1 wherein step c comprises the steps of:

i--dividing each resolution level which is en

coded into null and non-null regions;

ii--encoding said regions using binary tree decomposition;

iii--determining vector values describing the pixel values in said non-null regions; and

iv--quantizing said vector values.

13. The method in accordance with claim 12 wherein said vector values include DYADS, QUADS and DPCM.

14. The method in accordance with claim 1 wherein step a comprises the steps of:

i--forming a motion compensated difference image for each frame of said digital motion video signal;

ii--determining the value of a selected parameter associated with said motion compensated difference image;

iii--selecting a threshold value of said selected parameter; and

iv--selecting those frames from the sequence of frames having parameter values which exceed said predetermined thresholds.

15. A method of encoding a digital motion video signal comprising the steps of:

a--selecting at least one frame from a sequence of frames of said digital motion video signal;

b--encoding each selected frame as a still image including the steps of;

(1) resolving each selected frame into at least one lower level of resolution; and

(2) encoding at least one of the lower levels of resolution but not encoding the full resolution level; and

c--encoding each non-selected frame of said digital motion video signal using at least the full level of resolution.

16. The method in accordance with claim 15 in which step b(2) comprises, for each level of resolution of said selected frame which is encoded, the steps of:

i--forming a correction image comprising an array of pixels, by subtracting pixel values in a reference image from corresponding pixel values in that level of resolution of said selected frame;

ii--quantizing the pixel values in said correction image to form a quantized correction image; and

iii--encoding said quantized correction image.

17. The method in accordance with claim 16 in which step i for the lowest level of resolution of said selected frame includes the step of providing a reference frame comprising an array of pixels all of which have the same value.

18. The method in accordance with claim 17 in which step i for each level of resolution higher than the lowest level of resolution of said selected frame includes the steps of:

(1) forming a result image having a lower level of resolution by adding the pixel values of the quantized correction image at that lower level of resolution to the pixel values of the reference image at that lower level of resolution; and

(2) forming the reference image at the higher level of resolution by expanding the result image formed in step (1) to said higher level of resolution.

19. The method in accordance with claim 16 wherein step iii comprises the steps of dividing said quantized correction image into regions and generating region parameters descriptive of the position, size and fill data for each region.

20. The method in accordance with claim 19 wherein the step of dividing said quantized correction image into regions comprises dividing said image into null and non-null regions or into null and linear fill regions.

21. The method in accordance with claim 18 wherein said region descriptions are encoded using binary tree decomposition.

22. The method in accordance with claim 19 in which said non-null regions are encoded using vector quantization or discrete cosine transform technique.

23. The method in accordance with claim 22 wherein encoding using vector quantization comprises encoding using DYADS, QUADS, or DPCM.

24. The method in accordance with claim 15 in which step c comprises the steps of:

i--dividing each non-selected frame into regions and generating region parameters descriptive of the location and size of respective regions within said frame;

ii--for non-selected still frames or a first frame of a sequence of non-selected frames;

(1) determining absolute fill data describing respective regions;

(2) determining DPCM fill data describing respective regions for which absolute fill data cannot be determined;

(3) variable length encoding said absolute fill data, said DPCM fill data and corresponding region descriptive parameters;

(4) determining an estimate of decompression time from said variable length encoded absolute, DPCM fill data and corresponding region descriptive parameters;

(5) providing said variable length encoded fill data and corresponding region parameters if said estimate is less than a predetermined value, and if not, repeating steps (i) and (1)-(5) using a different tolerance in determining said absolute fill data; and

iii--for a succession of non-selected frames;

(1) comparing pixel values in each region with corresponding pixel values in a like sized region from a previously occurring frame to develop respective pixel differences;

(2) determining relative fill data descriptive of pixel differences for respective regions;

(3) determining DYAD fill data for respective regions in which relative fill data cannot be determined;

(4) determining DPCM fill data for respective regions in which relative fill data cannot be determined and for which DYAD fill data cannot be determined; and

(5) variable length coding said relative fill data, said DYAD fill data, said DPCM fill data and corresponding region parameters.

25. The method in accordance with claim 15 wherein step b(2) comprises the steps of:

i--dividing each resolution level which is encoded into null and non-null regions;

ii--encoding said regions using binary tree decomposition;

iii--determining vector values describing the pixel values in said non-null regions; and

iv--quantizing said vector values.

26. The method of claim 25 wherein said vector values include DYADS, QUADS and DPCM.

27. The method of claim 15 wherein step a comprises the steps of:

i--forming a motion compensated difference image for each frame of said digital motion video signal;

ii--determining the value of a selected parameter associated with said motion compensated difference image;

iii--selecting a threshold value of said selected parameter; and

iv--selecting those frames from the sequence of frames having parameter values which exceed said predetermined thresholds.

28. An apparatus for encoding a digital motion video signal comprising:

a--means for selecting at least one frame from a sequence of frames of said digital motion video signal;

b--means for resolving each selected frame into at least one lower level of resolution;

c--means for encoding at least one lower resolution level of each of said selected frames; and

d--means for encoding each non-selected frame of said digital motion video signal using only the full resolution level.

29. A method of decoding a digital motion video signal in which at least one selected frame from a sequence of frames has been resolved into at least one lower level of resolution which is then encoded, and in which each non-selected frame has been encoded at the full resolution level only, said method comprising the steps of:

a--decoding each level of resolution of each selected frame to form a correction image comprising an array of pixel values having that level of resolution;

b--forming a result image at each level of resolution by adding the pixel values in a reference image having that level of resolution to the pixel values of the expansion of the correction image to the same level of resolution; and

c--decoding said non-selected frames.

30. An apparatus for decoding a digital motion video signal comprising:

a--means for applying a digital motion video signal in which at least one selected frame from a sequence of frames has been resolved into at least one lower level of resolution which is then encoded, and in which each non-selected frame has been encoded at the " full resolution level only;

b--means for decoding each level of resolution of each selected frame to form a correction image comprising an array of pixel values having that level of resolution;

c--means for forming a result image at each level of resolution by adding the pixel values in a reference image having that level of resolution to the pixel values of the expansion of the correction image to the same level of resolution; and

d--means for decoding said non-selected frames.