Method for testing a motion vector

• US 8,457,424 B2
• Filed: 03/18/2009
• Issued: 06/04/2013
• Est. Priority Date: 03/18/2008
• Status: Expired due to Fees

• Current Assignee: Dynamic Data Technologies LLC
• Original Assignee: Entropic Communications Incorporated (Maxlinear Incorporated)
• Inventors: Hahn, Marko, Rieder, Peter, Tuschen, Christian, Scheffler, Gunter, Schu, Markus
• Primary Examiner(s): Do, Anh
• Application Number:
  US12/406,513
• Publication Number:
  US 20090238409A1
• Time in Patent Office: 1,539 Days
• Field of Search: 382/103, 382/107, 382/209, 382/216, 382/218, 382/236, 348/208.6, 348/208.4, 348/208.3, 348/333.01, 340/5.83, 375/240.12
• US Class Current: 382/236
• CPC Class Codes: G06T 2200/28 involving image processing hardware
• CPC Class Codes: G06T 2207/20021 Dividing image into blocks, subimages or windows
• CPC Class Codes: G06T 7/223 using block-matching

Abstract

A method for testing a motion vector is described, which has: provision of at least one item of motion information assigned to the image sequence; storing a first image section of the first image in a first buffer memory and storing a second image section of the second image in a second intermediate memory, whereby a position of the first image section in the first image and a position of the second image section in the second image have reciprocal offset, which is dependent on the at least one item of motion information; determining a first image block in the first image section and a second image block in a second image section using the motion vector; comparing the contents of the first and of the second image block.

Claims

1. In a device for testing a motion vector which has at least one item of motion information assigned to an image sequence, a method comprising:

storing a first image section of a first image of the image sequence in a first buffer memory;

storing a second image section of a second image of the image sequence in a second intermediate memory, a position of the first image section in the first image and a position of the second image section in the second image having a reciprocal offset which is dependent on the at least one item of motion information;

determining a first image block in the first image section and a second image block in the second image section using the motion vector; and

comparing the contents of the first and of the second image blocks.

2. The method as claimed in claim 1, wherein determining the motion information comprises:

ascertaining a motion vector field with several motion vectors which are assigned an image of the image sequence; and

determining the motion information depending on at least one motion vector of the motion vector field.

3. The method as claimed in claim 2, wherein the determining the motion information further comprises:

filtering several motion vectors of the motion vector field to ascertain a filter value; and

selecting the filter value as motion information.

4. The method as claimed in claim 2, wherein the motion information corresponds to an average value of several motion vectors of the motion vector field.

5. The method as claimed in claim 2, wherein the motion information corresponds to an average value of all motion vectors of the motion vector field.

6. The method as claimed in claim 2, wherein the motion information is represented by a shortest motion vector of the motion vector field.

7. The method as claimed in claim 2, wherein the motion information is represented by a longest motion vector of the motion vector field.

8. The method as claimed in claim 2, wherein the motion vector to be tested is part of a second motion vector field and wherein the motion information depends on the motion vector located in the first motion vector field at the same position as the motion vector to be tested or depends on motion vectors located in the first motion vector field adjacent to the position of the motion vector to be tested.

9. The method as claimed in claim 1, further comprising assigning an interpolation phase to the motion vector, wherein a position of the first and second image section inside the first and second image is dependent on the interpolation phase.

10. The method as claimed in claim 9, wherein the motion vector is part of a motion vector field, wherein the first and second image section are shifted relative to a basic position, wherein the basic position is dependent on the position of the motion vector within the motion vector field and wherein a shift of the first and second image section relative to the basic position is dependent on the interpolation phase.

11. The method as claimed in claim 10, wherein a position (u1, v1) of the first image section in the first image equals (u₀−

(1−

α

)·

Δ

u, v₀−

α

·

Δ

v)

and wherein a position of the second image section (u2, v2) in the second image equals (u₀−

(1−

α

)·

Δ

u, v₀−

(1−

α

)·

Δ

v), wherein;

α

designates the interpolation phase,

u₀ designates a first position component of the basic position,

v₀ designates a second position component of the basic position,

Δ

u designates a first component of the offset, and

Δ

v designates a second component of the offset.

12. The method as claimed in claim 10, wherein the motion vector field comprises motion vectors to which the image blocks of an intermediate image to be interpolated are assigned.

13. The method as claimed in claim 12, wherein the basic position is dependent on the position of an image block to be interpolated in the block grid to which the motion vector to be tested is assigned.

14. The method as claimed in claim 13, wherein the basic position is selected such that the image block is located at least approximately in the center of the image section if the offset is zero.

15. The method as claimed in claim 1, further comprising testing a first motion vector and a second motion vector, wherein the first and second image section have a first offset for testing the first motion vector and have a second offset for testing the second motion vector.

16. The method as claimed in claim 2, further comprising determining motion vector fields for original images of the image sequence and for intermediate images to be interpolated, wherein the offset is dependent on only at least one vector field, which was determined for an original image.

17. The method as claimed in claim 16, wherein the offset is reduced or set to zero in determining motion vector fields for intermediate images.