Inter-prediction method and image decoding device

1. A method for inter-prediction performed by a video decoding apparatus, the method comprising:

deriving a motion vector of a current block based on motion information decoded from a bitstream;

acquiring reference samples of a first reference block by using the motion vector, wherein reference samples of an external region located outside a reference picture among the first reference block are acquired from reference samples of a corresponding region corresponding to the external region within the reference picture; and

predicting the current block based on the acquired reference samples,

wherein the reference samples of the corresponding region within the reference picture are determined by shifting sample positions in the external region toward an inside of the reference picture by an offset determined based on a picture width.

2. The method of claim 1 , further comprising:

adjusting the motion vector to point a second reference block corresponding to the first reference block after the deriving of the motion vector,

wherein the corresponding region is a region included in the second reference block.

3. The method of claim 2 , wherein the adjusting of the motion vector is performed when the motion vector points an outside of the reference picture.

4. The method of claim 1 , further comprising:

indicating a directional relationship between the first reference block and a second reference block after the deriving of the motion vector,

wherein the second reference block corresponds to the first reference block and includes the corresponding region,

wherein the acquiring of the reference samples comprises:

acquiring reference samples from the corresponding region according to the indicated directional relationship.

5. The method of claim 4 , wherein the reference picture is in an equirectangular projection format for a 360-degree image,

wherein the indicating comprises:

when the external region is located outside an upper or lower end of the reference picture, indicating that the first reference block and the second reference block are opposite to each other in terms of the directional relationship with respect to a vertical direction of the reference picture,

wherein the acquiring of the reference samples comprises:

acquiring the reference samples of the corresponding region in a direction opposite to a direction of arrangement of the reference samples of the external region with respect to the vertical direction of the reference picture.

6. The method of claim 1 , wherein the corresponding region is a region obtained by shifting the external region in a horizontal direction by the offset derived from offset information contained in the bitstream.

7. The method of claim 6 , wherein, when the external region is located outside a left side of the reference picture, the corresponding region is obtained by shifting the external region in a rightward direction by the offset, and

wherein, when the external region is located outside a right side of the reference picture, the corresponding region is obtained by shifting the external region in a leftward direction by the offset.

8. A video decoding apparatus comprising:

a motion deriver configured to derive a motion vector of a current block based on motion information decoded from a bitstream;

a sample acquirer configured to acquire reference samples of a first reference block by using the motion vector, wherein reference samples of an external region located outside a reference picture among the first reference block are acquired from reference samples of a corresponding region corresponding to the external region within the reference picture; and

a prediction executor configured to predict the current block based on the acquired reference samples,

wherein the reference samples of the corresponding region within the reference picture are determined by shifting sample positions in the external region toward an inside of the reference picture by an offset determined based on a picture width.

9. The video decoding apparatus of claim 8 , further comprising:

a motion adjuster configured to adjust the motion vector to point a second reference block corresponding to the first reference block,

wherein the corresponding region is a region included in the second reference block.

10. The video decoding apparatus of claim 9 , wherein the motion adjuster adjusts the motion vector when the motion vector points an outside of the reference picture.

11. The video decoding apparatus of claim 8 , further comprising:

a relationship indicator configured to indicate a directional relationship between the first reference block and a second reference block after the deriving of the motion vector,

wherein the second reference block corresponds to the first reference block and includes the corresponding region,

wherein the sample acquirer acquires reference samples from the corresponding region according to the indicated directional relationship.

12. The video decoding apparatus of claim 11 , wherein the reference picture is in an equirectangular projection format for a 360-degree image,

wherein, when the external region is located outside an upper or lower end of the reference picture, the relationship indicator indicates that the first reference block and the second reference block are opposite to each other in terms of the directional relationship with respect to a vertical direction of the reference picture,

wherein the sample acquirer acquires the reference samples of the corresponding region in a direction opposite to a direction of arrangement of the reference samples of the external region with respect to the vertical direction of the reference picture.

13. The video decoding apparatus of claim 8 , wherein the corresponding region is a region obtained by shifting the external region in a horizontal direction by the offset derived from offset information contained in the bitstream.

14. The video decoding apparatus of claim 13 , wherein, when the external region is located outside a left side of the reference picture, the corresponding region is obtained by shifting the external region in a rightward direction by the offset, and

wherein, when the external region is located outside a right side of the reference picture, the corresponding region is obtained by shifting the external region in a leftward direction by the offset.

15. A method for inter-prediction performed by a video encoding apparatus, the method comprising:

determining a motion vector of a current block;

acquiring reference samples of a first reference block related to the motion vector, wherein reference samples of an external region located outside a reference picture among the first reference block are acquired from reference samples of a corresponding region corresponding to the external region within the reference picture; and

predicting the current block based on the acquired reference samples,

wherein the reference samples of the corresponding region within the reference picture are determined by shifting sample positions in the external region toward an inside of the reference picture by an offset determined based on a picture width.

16. A non-transitory recording medium storing a bitstream generated by an inter-prediction based video encoding method, the method comprising:

determining a motion vector of a current block;

acquiring reference samples of a first reference block related to the motion vector, wherein reference samples of an external region located outside a reference picture among the first reference block are acquired from reference samples of a corresponding region corresponding to the external region within the reference picture; and

predicting the current block based on the acquired reference samples,

wherein the reference samples of the corresponding region within the reference picture are determined by shifting sample positions in the external region toward an inside of the reference picture by an offset determined based on a picture width.