Motion compensation for interlaced digital video signals

• US 5,565,922 A
• Filed: 01/24/1996
• Issued: 10/15/1996
• Est. Priority Date: 04/11/1994
• Status: Expired due to Term

• Current Assignee: Google Technology Holdings LLC (Alphabet Inc.)
• Original Assignee: General Instrument Corporation (CommScope Holding Company, Inc.)
• Inventors: Krause, Edward A.
• Primary Examiner(s): Tran, Thai Q.
• Assistant Examiner(s): AU, AMELIA MEGAN
• Application Number:
  US08/590,867
• Time in Patent Office: 265 Days
• Field of Search: 348/699, 348/700, 348/701, 348/384, 348/390, 348/402, 348/413, 348/416, 348/409, 348/415, 348/400, 348/401
• US Class Current: 375/240.14
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 7/223 using block-matching
• CPC Class Codes: H04N 19/105 Selection of the reference unit for prediction within a chosen coding or prediction mode, e.g. adaptive choice of position and number of pixels used for prediction
• CPC Class Codes: H04N 19/112 according to a given display mode, e.g. for interlaced or progressive display mode
• CPC Class Codes: H04N 19/137 Motion inside a coding unit, e.g. average field, frame or block difference
• CPC Class Codes: H04N 19/42 characterised by implementation details or hardware specially adapted for video compression or decompression, e.g. dedicated software implementation H04N19/635 takes precedence
• CPC Class Codes: H04N 19/503 involving temporal prediction adaptive coding with adaptive selection between spatial and temporal predictive coding H04N19/107; adaptive coding with adaptive selection among a plurality of temporal predictive coding modes H04N19/109
• CPC Class Codes: H04N 19/577 Motion compensation with bidirectional frame interpolation, i.e. using B-pictures
• CPC Class Codes: H04N 19/61 in combination with predictive coding

Abstract

A method and apparatus are provided for motion compensated coding of interlaced digital video signals. A first prediction error signal is obtained indicative of the difference between a current portion from a field of a current video frame and a similarly situated first portion in the same field of a prior video frame. A second prediction error signal is obtained indicative of the difference between the current portion and a similarly situated second portion in an opposite field of the prior video frame. A third prediction error signal is obtained indicative of the difference between the current portion and an average of the first and second portions. One of the first, second and third prediction error signals is selected to represent the current portion. Motion vector information is provided to enable the current frame portion to which the selected prediction error signal corresponds to be reproduced by a decoder.

Claims

1. A method for motion compensated coding of digital video signals defining successive frames of interlaced video data including a current frame with even and odd fields of pixel data, and a prior frame with even and odd fields of pixel data, said method comprising the steps of:

(a) comparing portions of pixel data from said current frame even field to similarly situated portions in said prior frame even field to determine a first prediction error signal (PES) indicative of the difference therebetween, and a corresponding motion vector F1_U ;

(b) comparing portions of pixel data from said current frame even field to similarly situated portions in said prior frame odd field to determine a second PES indicative of the difference therebetween, and a corresponding motion vector G1_U ;

(c) comparing portions of pixel data from said current frame odd field to similarly situated portions in said prior frame even field to determine a third PES indicative of the difference therebetween, and a corresponding motion vector G2_U, said vectors G1_U and G2_U being unidirectional with a vector G;

(d) comparing portions of pixel data from said current frame odd field to similarly situated portions in said prior frame odd field to determine a fourth PES indicative of the difference therebetween, and a corresponding motion vector F2_U, said vectors F1_U and F2_U being unidirectional with a vector F; and

(e) selecting at least one of said first, second, third and fourth PESs for use in representing a corresponding current frame portion with a minimized error.

2. The method in accordance with claim 1, wherein said motion vectors F1_U and F2_U are the same as said vector F, and said motion vectors G1_U and G2_U are related to said vector G such that G=2·

G1_U =(2/3)G2_U.

3. The method in accordance with claim 1, comprising the further step of selecting one of:

(i) said first and fourth PESs;

(ii) said second and third PESs;

(iii) an average of said first and second PESs; and

(iv) an average of said third and fourth PESs;

to represent a corresponding current frame portion with a minimized error.

4. A method in accordance with claim 1, comprising the further step of:

providing specific ones of said motion vectors according to said selecting steps, thereby enabling reproduction of a corresponding current frame portion by a decoder.

5. The method in accordance with claim 1, wherein said selecting step minimizes at least one of:

(i) the amount of data necessary for transmission to enable a decoder to reproduce the current frame; and

(ii) the amount of data contained in a selected PES.

6. The method in accordance with claim 3, comprising the further step of providing:

(i) said vector F when said first and fourth PESs are selected;

(ii) said vector G when said second and third PESs are selected;

(iii) an average of said vectors F1_U and G1_U when said average of said first and second PESs is selected; and

(iv) an average of said vectors G2_U and F2_U when said average of said third and fourth PESs is selected;

to enable reproduction of a corresponding current frame portion by a decoder.

7. A method in accordance with claim 1, comprising the further step of encoding said selected PESs with data indicative of locations in said prior frame which are associated with said corresponding motion vectors.

8. A method in accordance with claim 1, wherein said current frame even and odd field pixel data portions comprise pairs of even and odd blocks of pixel data.

9. A method in accordance with claim 1, wherein said vectors F1_U and F2_U, and G1_U and G2_U, respectively, are found by selecting the PESs, which when summed, provide the smallest magnitude, and then finding the vectors for the selected PESs.

10. A method in accordance with claim 1, wherein said vectors F and G are the same vector.

11. A method in accordance with claim 5, wherein the encoded data comprises field identifier information and specific ones of said motion vectors according to said selecting steps.

12. A method in accordance with claim 1 comprising the further steps of:

(f) comparing portions of pixel data from said current frame even field to similarly situated portions in said prior frame even field to determine a fifth prediction error signal (PES) indicative of the difference therebetween, and a corresponding motion vector F1;

(g) comparing portions of pixel data from said current frame odd field to similarly situated portions in said prior frame odd field to determine a sixth PES indicative of the difference therebetween, and a corresponding motion vector F2;

(h) comparing portions of pixel data from said current frame even field to similarly situated portions in said prior frame odd field to determine a seventh PES indicative of the difference therebetween, and a corresponding motion vector G1;

(j) comparing portions of pixel data from said current frame odd field to similarly situated portions in one of said;

(i) prior frame even field; and

(ii) current frame even field;

to determine an eighth PES indicative of the difference therebetween, and a corresponding motion vector G2; and

(k) selecting at least one of said PESs for use in representing a corresponding current frame portion with a minimized error.

13. The method in accordance with claim 12, wherein said selecting step minimizes at least one of:

(i) the amount of data necessary for transmission to enable a decoder to reproduce the current frame; and

(ii) the amount of data contained in a selected PES.

14. The method in accordance with claim 12, wherein said motion vectors F1_U and F2_U are the same as said vector F_U, and said motion vectors G1_U and G2_U are related to said vector G_U such that G_U =2·

G1_U =(2/3)G2_U.

15. The method in accordance with claim 12, comprising the further steps of selecting one of a frame mode and a field mode to represent the current frame, wherein:

(1) if said frame mode is selected, selecting one of;

(i) said first and fourth PESs;

(ii) said second and third PESs;

(iii) an average of said first and second PESs; and

(iv) an average of said third and fourth PESs;

to represent a corresponding current frame portion with a minimized error; and

(m) if said field mode is selected;

(i) selecting one of said fifth, seventh, and an average of said fifth and seventh PESs to represent a corresponding current frame even field portion with a minimized error; and

(ii) selecting one of said sixth, eighth, and an average of said sixth and eighth PESs to represent a corresponding current frame odd field portion with a minimized error.

16. A method in accordance with claim 12, comprising the further step of:

providing specific ones of said motion vectors according to said selecting steps, thereby enabling reproduction of a corresponding current frame portion by a decoder.

17. A method in accordance with claim 12, wherein said vectors F1_U and F2_U, and G1_U and G2_U, respectively, are found by selecting the PESs, which when summed, provide the smallest magnitude, and then finding the vectors for the selected PESs.

18. A method in accordance with claim 12, wherein said current frame even and odd field pixel data portions of said steps (a), (b), (c) and (d) comprise pairs of even and odd blocks of pixel data.

19. A method in accordance with claim 12, wherein said current frame even and odd field pixel data portions of said steps (f), (g), (h) and said step (j), subpart (i) comprise pairs of even and odd blocks of pixel data.

20. A method in accordance with claim 12, wherein said current frame even and odd field pixel data portions of said steps (a), (b), (c), (d), (f), (g), (h) and said step (j), subpart (i) comprise pairs of even and odd blocks of pixel data.

21. A method for motion compensated coding of digital video signals defining successive frames of interlaced video data including a current frame with even and odd fields of pixel data, and a prior frame with even and odd fields of pixel data, said method comprising the steps of:

(a) comparing portions of pixel data from said current frame even field to similarly situated portions in said prior frame even field to determine a first prediction error signal (PES) indicative of the difference therebetween, and a corresponding motion vector F1;

(b) comparing portions of pixel data from said current frame even field to similarly situated portions in said prior frame odd field to determine a second PES indicative of the difference therebetween, and a corresponding motion vector G1;

(c) comparing portions of pixel data from said current frame odd field to similarly situated portions in said current frame even field to determine a third PES indicative of the difference therebetween, and a corresponding motion vector G2;

(d) comparing portions of pixel data from said current frame odd field to similarly situated portions in said prior frame odd field to determine a fourth PES indicative of the difference therebetween, and a corresponding motion vector F2;

(e) selecting a preferred even PES from one of said first and second PESs for use in representing a corresponding current frame even field portion; and

(f) selecting a preferred odd PES from one of said third and fourth PESs for use in representing a corresponding current frame odd field portion.

22. The method in accordance with claim 21 wherein said vectors F1 and F2 are unidirectional.

23. The method in accordance with claim 21, comprising the further steps of:

(a) selecting one of;

(i) said first PES;

(ii) said second PES; and

(iii) an average of said first and second PESs

to represent said current frame even field portion; and

(b) selecting one of;

(i) said third PES;

(ii) said fourth PES; and

(iii) an average of said third and fourth PESs to represent said current frame odd field portion.

24. A method in accordance with claim 21, comprising the further step of:

providing specific ones of said motion vectors according to said selecting steps, thereby enabling reproduction of a corresponding current frame portion by a decoder.

25. The method in accordance with claim 21, wherein said selecting step minimizes at least one of:

(i) the amount of data necessary for transmission to enable a decoder to reproduce the current frame; and

(ii) the amount of data contained in a selected PES.

26. A method for motion compensated coding of digital video signals defining successive frames of interlaced video data including a current frame with even and odd fields of pixel data, and a prior frame with even and odd fields of pixel data, said method comprising the steps of:

(a) comparing portions of current frame even field pixel data to similarly situated prior frame even field pixel data to determine a first prediction error signal (PES) indicative of the differences therebetween, and a corresponding motion vector F1;

(b) comparing portions of current frame odd field pixel data to similarly situated prior frame odd field pixel data to determine a second PES indicative of the differences therebetween, and a corresponding motion vector F2;

(c) comparing portions of current frame even field pixel data to similarly situated prior frame odd field pixel data to determine a third PES indicative of the differences therebetween, and a corresponding motion vector G1;

(d) comparing portions of current frame odd field pixel data to similarly situated prior frame even field pixel data to determine a fourth PES indicative of the differences therebetween, and a corresponding motion vector G2_A ;

(e) comparing portions of current frame odd field pixel data to similarly situated current frame even field pixel data to determine a fifth PES indicative of the differences therebetween, and a corresponding motion vector G2_B ;

(f) selecting a preferred even PES from one of said first and third PESs for use in representing a corresponding portion of a current frame even field; and

(g) selecting a preferred odd PES from one of said second, fourth and fifth PESs for use in representing a corresponding portion of a current frame odd field.

27. The method in accordance with claim 26, comprising the further steps of:

(a) selecting one of;

(i) said first PES;

(ii) said third PES; and

(iii) an average of said first and third PESs;

to represent said corresponding current frame even field portion; and

(b) selecting one of;

(i) said second PES;

(ii) said fourth PES;

(iii) said fifth PES;

(iv) an average of said second and fourth PESs;

(v) an average of said second and fifth PESs;

(vi) an average of said fourth and fifth PESs;

(vii) an average of said second, fourth and fifth PESs;

to represent said corresponding current frame odd field portion.

28. A method in accordance with claim 26, comprising the further step of:

providing specific ones of said motion vectors according to said selecting steps, thereby enabling reproduction of a corresponding current frame portion by a decoder.

29. The method in accordance with claim 26, wherein said selecting step minimizes at least one of:

(i) the amount of data necessary for transmission to enable a decoder to reproduce the current frame; and

(ii) the amount of data contained in a selected PES.

30. A method in accordance with claim 26, wherein said current frame even and odd field pixel data portions of said steps (a), (b), (c) and (d) comprise pairs of even and odd blocks of pixel data.

31. Encoder apparatus for motion compensated coding of digital video signals defining successive frames of interlaced video data including current and prior frames with both even and odd fields of pixel data, said apparatus comprising:

(a) means for subdividing said current and prior frames into portions of pixel data;

(b) means for storing pixel data from successive fields for use in comparing portions of pixel data from said current frame even field to similarly situated portions in said prior frame even and odd fields, and for comparing portions of pixel data from said current frame odd field to similarly situated portions in at least one of;

(i) said prior frame even and odd fields; and

(ii) said prior frame even and odd fields and said current frame even field; and

(iii) said current frame even field;

to determine corresponding prediction error signals (PESs) indicative of the differences therebetween, and corresponding motion vectors; and

(c) means for selecting particular ones of said PESs for use in representing corresponding current frame even and odd field portions.

32. The apparatus in accordance with claim 31, wherein a first set of unidirectional motion vectors corresponds to selected same-field PESs, and a second set of unidirectional motion vectors correspond to selected opposite-field PESs.

33. The apparatus in accordance with claim 31, further comprising:

(a) means for averaging selected ones of said PESs to obtain averaged even and odd field PESs;

(b) means for selecting particular ones of said averaged and unaveraged PESs to represent said current frame even and odd field portions.

34. The apparatus in accordance with claim 31, further comprising:

means for providing specific ones of said motion vectors according to said selecting steps, thereby enabling reproduction of a current frame portion by a decoder.

35. The apparatus in accordance with claim 31, further comprising:

means for scaling selected ones of said motion vectors associated with said opposite field PESs to account for temporal differences resulting from the use of an opposite field comparison instead of a same field comparison.

36. Decoder apparatus for motion compensated interlaced digital video signals which operates in at least one of a plurality of operating modes, said apparatus comprising:

(a) memory means for storing odd and even field pixel data for a first video frame;

(b) means for receiving motion compensated pixel data from odd and even field portions of a second video frame together with motion vector and motion compensation mode information for said portions, said motion vector information identifying pixel data in said first video frame such that said second video frame can be reconstructed using said received motion compensated pixel data;

(c) motion compensator means responsive to received motion vector information for retrieving pixel data from designated fields of said first video frame; and

(d) means responsive to said motion compensation mode information for combining said received motion compensated pixel data from said second frame with said retrieved pixel data from said first frame in order to reconstruct said second frame;

wherein;

(i) in a first of said operating modes, both even and odd field pixel data from said first frame is used to reconstruct the second frame even and odd fields based on same-field motion vectors that are unidirectional;

(ii) in a second of said operating modes, both even and odd field pixel data from said first frame are used to reconstruct the second frame even and odd fields based on same-field motion vectors that are not unidirectional; and

(iii) in a third of said operating modes, first frame even and odd field pixel data is used to reconstruct the second frame even field, and subsequently, the first frame even and odd field pixel data and second frame even field pixel data are used to reconstruct the second frame odd field.

37. Decoder apparatus for motion compensated interlaced digital video signals defining successive frames of video data including a current frame with even and odd fields of pixel data, and a prior frame with even and odd fields of pixel data, said apparatus comprising:

means for receiving a first motion vector F for use in reconstructing corresponding portions of current frame even and odd fields of pixel data from prior frame same-field pixel data;

means for receiving a second motion vector G for use in reconstructing corresponding portions of current frame even and odd fields of pixel data from prior frame opposite-field pixel data;

wherein said vector G is scaled as required to account for temporal differences resulting from the use of opposite field data instead of same field data.

38. The apparatus of claim 37, wherein said vectors F and G are unidirectional.

39. Decoder apparatus for motion compensated interlaced digital video signals defining successive frames of video data including a current frame with even and odd fields of pixel data, and a prior frame with even and odd fields of pixel data, said apparatus comprising:

means for receiving a motion vector for use in reconstructing corresponding portions of current frame even and odd fields of pixel data from prior frame same-field and opposite field pixel data;

wherein said vector is scaled as required to account for temporal differences resulting from the use of opposite field data instead of same field data.