METHOD OF PREDICTING MOTION VECTORS IN VIDEO CODEC IN WHICH MULTIPLE REFERENCES ARE ALLOWED, AND MOTION VECTOR ENCODING/DECODING APPARATUS USING THE SAME

• US 20140254685A1
• Filed: 05/20/2014
• Published: 09/11/2014
• Est. Priority Date: 05/26/2010
• Status: Active Grant

• Current Assignee: Hangzhou Hikvision Digital Technology Company Limited
• Original Assignee: Newratek, Inc.
• Inventors: Jeong, Je Chang, Jin, Soon Jong, Park, Sang Jun, Lee, Hyuk
• Granted Patent: US 9,344,738 B2
• Time in Patent Office: Days
• US Class Current: 375/240.16
• CPC Class Codes: H04N 19/176 the region being a block, e.g. a macroblock
• CPC Class Codes: H04N 19/513 Processing of motion vectors
• CPC Class Codes: H04N 19/52 by predictive encoding
• CPC Class Codes: H04N 19/527 Global motion vector estimation
• CPC Class Codes: H04N 19/55 Motion estimation with spatial constraints, e.g. at image or region borders
• CPC Class Codes: H04N 19/573 Motion compensation with multiple frame prediction using two or more reference frames in a given prediction direction

Abstract

Initial value is selected when motion estimation is performed using predicted motion vector in video codec in which multiple references are allowed, and coding amount of motion vectors is reduced using predicted motion vector. Motion vector encoding apparatus includes predicted motion vector generator that generates adjacent motion vectors of current block with respect to each of reference pictures and generates each predicted motion vector of the current block from the generated adjacent motion vectors, motion estimator that determines final motion vector of the current block and final reference picture corresponding to the final motion vector with regard to search regions of the reference pictures, differential unit that obtains difference between the final motion vector of the current block and final predicted motion vector corresponding to the determined final reference picture among the predicted motion vectors, and unit for encoding information about the final reference picture and the difference.

Claims

1. A method of a decoding device to decode a current block in a current picture, comprising:

obtaining, by the decoding device, a motion vector of a first neighboring block of the current block, wherein the first neighboring block is in the current picture;

interpolating, by the decoding device, the motion vector of the first neighboring block in the current picture based on a distance between a reference picture of the first neighboring block in the current picture and the current picture and a distance between a reference picture of the current block in the current picture and the current picture to generate a first adjacent motion vector for predicting a motion vector of the current block;

obtaining, by the decoding device, a motion vector of a second neighboring block of the current block, wherein the second neighboring block is in the current picture and the second neighboring block is different from the first neighboring block;

interpolating, by the decoding device, the motion vector of the second neighboring block in the current picture based on a distance between a reference picture of the second neighboring block in the current picture and the current picture and a distance between a reference picture of the current block in the current picture and the current picture to generate a second adjacent motion vector for predicting the motion vector of the current block;

predicting, by the decoding device, the motion vector of the current block based on the first adjacent motion vector and the second adjacent motion vector to generate a predicted motion vector of the current block;

performing, by the decoding device, a motion compensating by using the predicted motion vector of the current block to generate a motion compensated block of the current block; and

adding, by the decoding device, a residual block of the current block and the motion compensated block of the current block to reconstruct the current block.

2. The method of claim 1, wherein the first neighboring block is formed in a rectangular shape by asymmetric partition.

3. The method of claim 1, wherein the second neighboring block is formed in a rectangular shape by asymmetric partition.

4. The method of claim 1, wherein the method is applied to a video coding scheme in which multiple references are allowed.

5. The method of claim 1, wherein interpolating the motion vector of the first neighboring block comprises:

interpolating the motion vector of the first neighboring block in the current picture based on the distance between the reference picture of the first neighboring block in the current picture and the current picture and the distance between the reference picture of the current block in the current picture and the current picture to generate the first adjacent motion vector for predicting the motion vector of the current block, when the distance between the reference picture of the first neighboring block and the current picture is equal to or greater than the distance between the reference picture of the current block and the current picture.

6. The method of claim 5, further comprising:

continuously tracing a location to which the motion vector of the first neighboring block points to generate the first adjacent motion vector for predicting the motion vector of the current block, and

the first adjacent motion vector is obtained by a displacement between a region centering on the traced location in the reference picture of the current block and the first neighboring block.

7. The method of claim 1, wherein the first neighboring block is one of a left block, an upper block, and a right upper block of the current block.

8. The method of claim 1, wherein the second neighboring block is one of a left block, an upper block, and a right upper block of the current block.

9. The method of claim 1, wherein the first neighboring block is a left block of the current block, and the second neighboring block is one of an upper block, and a right upper block of the current block.

10. A decoding device for decoding a current block in a current picture, comprising:

a prediction motion vector generator configured to interpolate a motion vector of the first neighboring block of the current block based on a distance between a reference picture of the first neighboring block in the current picture and the current picture and a distance between a reference picture of the current block in the current picture and the current picture to generate a first adjacent motion vector for predicting a motion vector of the current block,

interpolate a motion vector of the second neighboring block of the current block based on a distance between a reference picture of the second neighboring block in the current picture and the current picture and a distance between a reference picture of the current block in the current picture and the current picture to generate a second adjacent motion vector for predicting the motion vector of the current block, and

predict the motion vector of the current block based on the first adjacent motion vector and the second adjacent motion vector to generate a predicted motion vector of the current block, wherein the first neighboring block is in the current picture, the second neighboring block is in the current picture, and the second neighboring block is different from the first neighboring block;

a motion compensator configured to generate a motion compensated block of the current block by using the predicted motion vector of the current block; and

an adder configured to add a residual block of the current block and the motion compensated block of the current block to reconstruct the current block.

11. The decoding device of claim 10, wherein the first neighboring block is formed in a rectangular shape by asymmetric partition.

12. The decoding device of claim 10, wherein the second neighboring block is formed in a rectangular shape by asymmetric partition.

13. The decoding device of claim 10, wherein the method is applied to a video coding scheme in which multiple references are allowed.

14. The decoding device of claim 10, wherein the prediction motion vector generator interpolates the motion vector of the first neighboring block in the current picture based on the distance between the reference picture of the first neighboring block in the current picture and the current picture and the distance between the reference picture of the current block in the current picture and the current picture to generate the first adjacent motion vector for predicting the motion vector of the current block, when the distance between the reference picture of the first neighboring block and the current picture is equal to or greater than the distance between the reference picture of the current block and the current picture.

15. The decoding device of claim 14, wherein the prediction motion vector generator continuously traces a location to which the motion vector of the first neighboring block points to generate the first adjacent motion vector for predicting the motion vector of the current block, and the first adjacent motion vector is obtained by a displacement between a region centering on the traced location in the reference picture of the current block and the first neighboring block.

16. The decoding device of claim 10, wherein the first neighboring block is one of a left block, an upper block, and a right upper block of the current block.

17. The decoding device of claim 10, wherein the second neighboring block is one of a left block, an upper block, and a right upper block of the current block.

18. The decoding device of claim 10, wherein the first neighboring block is a left block of the current block, and the second neighboring block is one of an upper block, and a right upper block of the current block.