Motion estimation employing adaptive spatial update vectors

• US 7,920,627 B2
• Filed: 02/28/2005
• Issued: 04/05/2011
• Est. Priority Date: 04/30/2004
• Status: Expired due to Fees

• Current Assignee: Panasonic Corporation (Panasonic Holdings Corporation)
• Original Assignee: Panasonic Corporation (Panasonic Holdings Corporation)
• Inventors: Grundmeyer, Michael, Hubrich, Ralf
• Primary Examiner(s): Czekaj; David
• Assistant Examiner(s): PONTIUS, JAMES M
• Application Number:
  US11/067,649
• Publication Number:
  US 20050243927A1
• Time in Patent Office: 2,227 Days
• Field of Search: 375/240.16
• US Class Current: 375/240.16
• CPC Class Codes: H04N 5/145 Movement estimation for video coding H04N19/51

Abstract

A method for motion estimation and for motion compensated interpolation. By taking a source of video data into account, a size of update vectors to be employed for determining spatial prediction vectors is set in accordance with a detected source mode. By selecting an appropriate update vector size, accuracy of predicted motion and consequently, a picture quality of motion compensated interpolated images can be increased considerably.

Claims

1. A method for determining a motion vector for a block of a current image in a sequence of video images, the sequence of video images including image data of a film mode and a video mode, each video image of the sequence of video images being divided into a plurality of blocks, the method comprising:

determining, using a motion estimator, a motion vector for a current block by combining a previously determined motion vector and a predetermined update vector; and

setting a size of the predetermined update vector used for the determining of the motion vector for the current block, the size of the predetermined update vector being set to be larger when the image data of the current block is of the film mode than the size of the predetermined update vector when the image data of the current block is of the video mode.

2. The method according to claim 1, wherein a determination of whether the image data is of the film mode or the video mode is based on a detection of a motion picture to video data conversion pattern in the sequence of video images.

3. The method according to claim 2, wherein the motion picture to video data conversion pattern is a 2:

2 conversion pattern or a 3;

2 conversion pattern.

4. The method according to claim 1, wherein a determination of whether the image data is of the film mode or the video mode is determined on an image-by-image basis, in particular on a per field or frame basis.

5. The method according to claim 1, wherein a determination of whether the image data of the current block is of the film mode or the video mode is performed individually for the current block, independently of other blocks of the sequence of video images.

6. The method according to claim 1, wherein, when the image data is of the film mode, the size of the predetermined update vector is set between 1.5 and 2.5 times as large, preferably twice as large, in comparison to the size of the predetermined update vector when the image data is of the video mode.

7. The method according to claim 1, wherein the predetermined update vector is selected from a plurality of predetermined update vectors.

8. The method according to claim 1, further comprising:

selecting the motion vector for the current block from a plurality of candidate motion vectors including the motion vector determined by combining the previously determined motion vector and the predetermined update vector; and

assigning the selected motion vector to the current block.

9. The method according to claim 8, wherein the selecting comprises:

calculating an error value for each of the plurality of candidate motion vectors; and

selecting the motion vector of the plurality of candidate motion vectors that has a smallest error value.

10. The method according to claim 8, wherein the plurality of candidate motion vectors includes a plurality of combinations of a given previously determined motion vector with different predetermined update vectors from a plurality of predetermined update vectors.

11. The method according to claim 1, wherein the previously determined motion vector includes at least one of the following motion vectors:

a zero motion vector pointing to an identical block position of the current block;

a motion vector determined for an adjacent block in the current image; and

a motion vector determined for a block in a previous image.

12. The method according to claim 8, wherein the plurality of candidate motion vectors includes at least one of the following motion vectors:

a zero motion vector pointing to an identical block position of the current block;

a motion vector determined for an adjacent block in the current image; and

a motion vector determined for a block in a previous image.

13. A method for encoding a sequence of video images including motion compensation employing a motion vector determining method in accordance with claim 1.

14. A method for interpolating a sequence of video images including motion compensation employing a motion vector determining method in accordance with claim 1.

15. A method for converting a field- or frame-rate of a video sequence by employing motion compensation in accordance with claim 14.

31. The method according to claim 2, wherein a determination of whether the image data is of the film mode or the video mode is determined on an image-by-image basis, in particular on a per field or frame basis.

32. The method according to claim 2, wherein a determination of whether the image data is of the film mode or the video mode is determined on a block-by-block basis.

16. A motion estimator for determining a motion vector for a block of a current image in a sequence of video images, the sequence of video images including image data of a film mode and a video mode, each video image of the sequence of video images being divided into a plurality of blocks,

wherein the motion estimator determines a motion vector for a current block by combining a previously determined motion vector and a predetermined update vector, and

wherein the motion estimator comprises:

a film mode detector determining whether the image data of the current block is of the film mode or the video mode, and

adjusting means for setting a size of the predetermined update vector used for determining the motion vector for the current block, the size of the predetermined update vector being set to be larger when the image data of the current block is of the film mode than the size of the predetermined update vector when the image data of the current block is of the video mode.

17. The motion estimator according to claim 16, wherein the film mode detector determines whether the image data is of the film mode or the video mode based on a detection of a motion picture to video data conversion pattern in the sequence of video images.

18. The motion estimator according to claim 17, wherein the motion picture to video data conversion pattern is a 2:

2 conversion pattern or a 3;

2 conversion pattern.

19. The motion estimator according to claim 16, wherein the film mode detector determines whether the image data is of the film mode or the video mode on an image-by-image basis, in particular on a per field or frame basis.

20. The motion estimator according to claim 16, wherein the film mode detector determines whether the image data of the current block is of the film mode or the video mode, individually for the current block, independently of other blocks of the sequence of video images.

21. The motion estimator according to claim 16, wherein, when the image data is of the film mode, the adjusting means sets the size of the predetermined update vector 1.5 to 2.5 times as large, preferably twice as large, in comparison to the size of the predetermined update vector when the image data is of the video mode.

22. The motion estimator according to claim 16, wherein the adjusting means comprises a memory and selects the predetermined update vector from a plurality of pre-stored update vectors.

23. The motion estimator according to claim 16, further comprising a selector for selecting the motion vector for the current block from a plurality of candidate motion vectors including the motion vector determined by combining the previously determined motion vector and the predetermined update vector.

24. The motion estimator according to claim 23, wherein the selector comprises:

a processing unit for calculating an error value for each of the plurality of candidate motion vectors; and

a comparator for selecting the motion vector of the plurality of candidate motion vectors that has a smallest error value.

25. The motion estimator according to claim 23, wherein the plurality of candidate motion vectors includes a plurality of combinations of a given previously determined motion vector with different predetermined update vectors from a plurality of predetermined update vectors.

26. The motion estimator according to claim 16, wherein the previously determined motion vector includes at least one of the following motion vectors:

a zero motion vector pointing to an identical block position of the current block;

a motion vector determined for an adjacent block in the current image; and

a motion vector determined for a block in a previous image.

27. The motion estimator according to claim 23, wherein the plurality of candidate motion vectors includes at least one of the following motion vectors:

a zero motion vector pointing to an identical block position of the current block;

a motion vector determined for an adjacent block in the current image; and

a motion vector determined for a block in a previous image.

28. An encoder for encoding a sequence of video images including motion compensation employing a motion estimator in accordance with claim 16.

29. An interpolator for interpolating a sequence of video images including motion compensation employing a motion estimator in accordance with claim 16.

30. A converter for converting a field- or frame-rate of a video sequence by employing motion compensation in accordance with claim 29.

33. The motion estimator according to claim 17, wherein the film mode detector determines whether the image data is of the film mode or the video mode on an image-by-image basis, in particular on a per field or frame basis.

34. The motion estimator according to claim 17, wherein the film mode detector determines whether the image data is of the film mode or the video mode on a block-by-block basis.