Method and apparatus for inter prediction in video processing system
Disclosed is an inter prediction method which includes deriving control points (CPs) for the current block, wherein the CPs include a first CP and a second CP, deriving a first motion vector predictor (MVP) for the first CP and a second MVP for the second CP based on neighboring blocks of the current block, decoding a first motion vector difference (MVD) for the first CP, decoding a difference of two MVDs (DMVD) for the second CP, deriving a first motion vector (MV) for the first CP based on the first MVP and the first MVD, deriving a second MV for the second CP based on the second MVP and the DMVD for the second CP, and generating a predicted block for the current block based on the first MV and the second MV.
1. An image decoding method performed by a decoding apparatus, the method comprising:
deriving a first motion vector predictor (MVP) for a first control point (CP), a second MVP for a second CP, and a third MVP for a third CP of a current block based on neighboring blocks of the current block, wherein the first CP is located at a top-left position of the current block, the second CP is located at a top-right position of the current block, and the third CP is located at a bottom-left position of the current block;
decoding a first motion vector difference (MVD) for the first CP;
decoding a difference of two MVDs (DMVD) for the second CP;
decoding a DMVD for the third CP;
deriving a first motion vector (MV) for the first CP based on the first MVP and the first MVD;
deriving a second MV for the second CP based on the second MVP and the DMVD for the second CP;
deriving a third MV for the third CP based on the third MVP and the DMVD for the third CP;
generating prediction samples for the current block based on the first MV, the second MV, and the third MV; and
generating reconstructed samples based on the prediction samples for the current block,
wherein the DMVD for the second CP is for a difference between a second MVD for the second CP and the first MVD,
wherein the DMVD for the third CP is for a difference between a third MVD for the third CP and the first MVD,
wherein the deriving the second MV for the second CP comprises:
deriving the second MVD for the second CP based on the first MVD and the DMVD for the second CP; and
deriving the second MV for the second CP based on the second MVP and the second MVD which is derived based on the DMVD for the second CP, and
wherein the deriving the third MV for the third CP comprises:
deriving the third MVD for the third CP based on the first MVD and the DMVD for the third CP; and
deriving the third MV for the third CP based on the third MVP and the third MVD which is derived based on the DMVD for the third CP.
2. The method of claim 1 , wherein the first MVP is derived based on neighboring blocks of the first CP.
3. The method of claim 1 , wherein the second MVP is derived based on neighboring blocks of the second CP.
4. The method of claim 1 , wherein the third MVP is derived based on neighboring blocks of the third CP.
5. The method of claim 1 , wherein the third MVP is derived based on the first MVP and the second MVP.
6. The method of claim 5 , wherein the third MVP is derived based on the following equation,
{
mvp
2
x
=
mvp
0
x
-
h
*
(
mvp
1
y
-
mvp
0
y
)
w
mvp
2
y
=
mvp
0
y
+
h
*
(
mvp
1
x
-
mvp
0
x
)
w
herein mvp 0x and mvp 0y are x and y components of the first MVP, mvp 1x and mvp 1y are x and y components of the second MVP, mvp 2x and mvp 2y are x and y components of the third MVP, and h and w are the height and width of the current block.
7. The method of claim 1 , wherein the DMVD for the second CP is decoded based on a case that a median value derived based on the first MVP, the second MVP, and the third MVP is same as the first MVP.
8. An image encoding method performed by an encoding apparatus, the method comprising:
deriving a first motion vector predictor (MVP) for a first control point (CP), a second MVP for a second CP, and a third MVP for a third CP of a current block based on neighboring blocks of the current block, wherein the first CP is located at a top-left position of the current block, the second CP is located at a top-right position of the current block, and the third CP is located at a bottom-left position of the current block;
deriving a first motion vector difference (MVD) for the first CP;
deriving a difference of two MVDs (DMVD) for the second CP;
deriving a DMVD for the third CP; and
encoding image information including information on the first MVD, information on the DMVD for the second CP, and information on the DMVD for the third CP to output a bitstream,
wherein the DMVD for the second CP is for a difference between a second MVD for the second CP and the first MVD,
wherein the DMVD for the third CP is for a difference between a third MVD for the third CP and the first MVD,
wherein the deriving the DMVD for the second CP comprises:
deriving the second MVD for the second CP based on a second MV for the second CP and the second MVP for the second CP; and
deriving the DMVD for the second CP based on the second MVD and the first MVD,
wherein the deriving the DMVD for the third CP comprises:
deriving the third MVD for the third CP based on a third MV for the third CP and the third MVP for the third CP; and
deriving the DMVD for the third CP based on the third MVD and the first MVD.
9. The method of claim 8 , wherein the first MVP is derived based on neighboring blocks of the first CP.
10. The method of claim 8 , wherein the second MVP is derived based on neighboring blocks of the second CP.
11. The method of claim 8 , wherein the third MVP is derived based on neighboring blocks of the third CP.
12. The method of claim 8 , wherein the third MVP is derived based on the first MVP and the second MVP.
13. The method of claim 12 , wherein the third MVP is derived based on the following equation,
{
mvp
2
x
=
mvp
0
x
-
h
*
(
mvp
1
y
-
mvp
0
y
)
w
mvp
2
y
=
mvp
0
y
+
h
*
(
mvp
1
x
-
mvp
0
x
)
w
herein mvp 0y and mvp 0y are x and y components of the first MVP, mvp 1x and mvp 1y are x and y components of the second MVP, mvp 2x and mvp 2y are x and y components of the third MVP, and h and w are the height and width of the current block.
14. The method of claim 8 , wherein the DMVD for the second CP is coded based on a case that a median value derived based on the first MVP, the second MVP, and the third MVP is same as the first MVP.
15. A non-transitory computer readable storage medium storing a bitstream generated by a method, the method comprising:
deriving a first motion vector predictor (MVP) for a first control point (CP), a second MVP for a second CP, and a third MVP for a third CP of a current block based on neighboring blocks of the current block, wherein the first CP is located at a top-left position of the current block, the second CP is located at a top-right position of the current block, and the third CP is located at a bottom-left position of the current block;
deriving a first motion vector difference (MVD) for the first CP;
deriving a difference of two MVDs (DMVD) for the second CP;
deriving a DMVD for the third CP; and
encoding image information including information on the first MVD, information on the DMVD for the second CP, and information on the DMVD for the third CP to output a bitstream,
wherein the DMVD for the second CP is for a difference between a second MVD for the second CP and the first MVD,
wherein the DMVD for the third CP is for a difference between a third MVD for the third CP and the first MVD,
wherein the deriving the DMVD for the second CP comprises:
deriving the second MVD for the second CP based on a second MV for the second CP and the second MVP for the second CP; and
deriving the DMVD for the second CP based on the second MVD and the first MVD,
wherein the deriving the DMVD for the third CP comprises:
deriving the third MVD for the third CP based on a third MV for the third CP and the third MVP for the third CP; and
deriving the DMVD for the third CP based on the third MVD and the first MVD.