Video signal processing method and device

1. A method of decoding an image, the method comprising:

generating a merge candidate list of a current block;

selecting one among a plurality of merge candidates included in the merge candidate list;

determining whether to modify a motion vector derived from the selected merge candidate;

determining a motion refinement vector of the current block, when it is determined to modify the motion vector; and

determining a modified motion vector of the current block of based on the motion vector and the motion refinement vector,

wherein the motion refinement vector is determined based on information decoded from a bitstream,

wherein the information includes magnitude information of the motion refinement vector and direction information of the motion refinement vector,

wherein the magnitude information indicates any one of multiple magnitude candidates, and the direction information indicates any one of multiple direction candidates,

wherein the one among the plurality of the merge candidates is selected based on index information decoded from the bitstream, and

wherein when it is determined to modify the motion vector, a number of merge candidates selectable by the index information is less than when it is determined not to modify the motion vector.

2. The method of claim 1 , wherein when it is determined to modify the motion vector, the number of merge candidates selectable by the index information is 2 .

3. The method of claim 1 , wherein a maximum value among the multiple magnitude candidates is varied according to motion vector precision information.

4. The method of claim 3 , wherein when the motion vector precision information indicates a fractional pel, the maximum value among the multiple magnitude candidates does not exceed a pre-defined value.

5. The method of claim 1 , wherein a merge refinement vector is determined based on a result of searching a search region including a position indicated by the motion vector, for a region in which a rate-distortion cost value is the lowest.

6. The method of claim 5 , wherein a vector from the motion vector to a position of the region in which the rate-distortion cost value is the lowest is determined as the merge refinement vector.

7. The method of claim 1 , wherein the modified motion vector is a sum of the motion vector and the motion refinement vector.

8. A method of encoding an image, the method comprising:

generating a merge candidate list of a current block;

selecting one among a plurality of merge candidates included in the merge candidate list;

determining whether to modify a motion vector derived from the selected merge candidate;

determining a motion refinement vector of the current block; and

determining a modified motion vector of the current block based on the motion vector and the motion refinement vector,

wherein information for determining the motion refinement vector is encoded to a bitstream,

wherein the information includes magnitude information of the motion refinement vector and direction information of the motion refinement vector,

wherein the magnitude information indicates any one of multiple magnitude candidates and the direction information indicates any one of multiple direction candidates,

wherein index information for selecting the one among the plurality of the merge candidates is encoded to the bitstream, and

wherein when it is determined to modify the motion vector, a number of merge candidates selectable by the index information is less than when it is determined not to modify the motion vector.

9. The method of claim 8 , wherein when it is determined to modify the motion vector, the number of merge candidates selectable by the index information is 2 .

10. The method of claim 8 , wherein a maximum value among the multiple magnitude candidates is varied according to motion vector precision information.

11. The method of claim 9 , wherein when the motion vector precision information indicates a fractional pel, the maximum value among the multiple magnitude candidates does not exceed a pre-defined value.

12. The method of claim 8 , wherein a merge refinement vector is determined based on a result of searching a search region including a position indicated by the motion vector, for a region in which a rate-distortion cost value is the lowest.

13. The method of claim 12 , wherein a vector from the motion vector to a position of the region in which the rate-distortion cost value is the lowest is determined as the merge refinement vector.

14. The method of claim 8 , wherein the modified motion vector is a sum of the motion vector and the motion refinement vector.

15. A non-transitory computer-readable storage medium having instructions stored thereon that, when executed, cause one or more processor to:

generating a merge candidate list of a current block;

selecting one among a plurality of merge candidates included in the merge candidate list;

determining whether to modify a motion vector derived from the selected merge candidate;

determining a motion refinement vector of the current block, when it is determined to modify the motion vector; and

determining a modified motion vector of the current block based on the motion vector and the motion refinement vector,

wherein the motion refinement vector is determined based on information decoded from a bitstream,

wherein the information includes magnitude information of the motion refinement vector and direction information of the motion refinement vector,

wherein the magnitude information indicates any one of multiple magnitude candidates, and the direction information indicates any one of multiple direction candidates,

wherein the one among the plurality of the merge candidates is selected based on index information decoded from the bitstream, and

wherein when it is determined to modify the motion vector, a number of merge candidates selectable by the index information is less than when it is determined not to modify the motion vector.