Image processing apparatus and method
The present disclosure relates to an image processing apparatus and an image processing method capable of improving process efficiency through a parallel process. In a case of corresponding PU 1 , there is a use of motion vector information pieces of B 1 , C 1 , and E 1 which are adjacent to corresponding PU 1 in the following positional relationship, and A 0 and D 0 which are adjacent to PU 0 located on corresponding PU 1 in the CU in the following positional relationship. In other words, in relation to corresponding PU 1 , a PU corresponding to A 1 is PU 0 , and thus A 0 is set instead of A 1 as an adjacent region of corresponding PU 1 . The present disclosure is applicable to, for example, an image processing apparatus.
1. An image processing apparatus comprising:
an adjacent region setting unit that sets spatially adjacent regions on the basis of positions of prediction regions in a coding region so that generation of spatial motion vectors of the plurality of prediction regions is performed in parallel when a spatial prediction motion vector is generated among prediction motion vectors used to decode motion vectors of the plurality of prediction regions forming the coding region of an image;
a prediction motion vector generation unit that generates a spatial prediction vector of the prediction region by using the motion vector of the spatially adjacent region set by the adjacent region setting unit; and
a motion vector decoding unit that decodes a motion vector of the prediction region by using the prediction motion vector of the prediction region,
wherein, in a case where a first prediction region is located on a right or lower side in the coding region, the adjacent region setting unit sets a first adjacent region of a second prediction region instead of the first adjacent region serving as the second prediction region located on a left or upper side in the coding region among adjacent regions of the first prediction region, and
wherein the adjacent region setting unit, the prediction motion vector generation unit, and the motion vector decoding unit are each implemented via at least one processor.
2. The image processing apparatus according to claim 1 ,
wherein, in a case where the first prediction region is located on the right or lower side in the coding region, the adjacent region setting unit sets a second adjacent region of the second prediction region instead of the second adjacent region adjacent to the first prediction region among the adjacent regions of the first prediction region.
3. The image processing apparatus according to claim 2 ,
wherein, in a case where the coding region is split into prediction regions of 2N×N, the first prediction region is a prediction region located on the lower side in the coding region, the second prediction region is a prediction region located on the upper side in the coding region, the first adjacent region is an adjacent region adjacent to an upper side of the prediction region, and the second adjacent region is an adjacent region adjacent to an upper right side of the prediction region.
4. The image processing apparatus according to claim 2 ,
wherein, in a case where the coding region is split into prediction regions of N×2N, the first prediction region is a prediction region located on the right side in the coding region, the second prediction region is a prediction region located on the left side in the coding region, the first adjacent region is an adjacent region adjacent to a left side of the prediction region, and the second adjacent region is an adjacent region adjacent to a lower left side of the prediction region.
5. The image processing apparatus according to claim 2 ,
wherein, in a case where the first prediction region is located on the right side and the lower side in the coding region, the adjacent region setting unit sets a first adjacent region of a second prediction region instead of the first adjacent region serving as the second prediction region adjacent to the upper left side in the coding region among adjacent regions of the first prediction region.
6. The image processing apparatus according to claim 5 ,
wherein, in a case where the coding region is split into prediction regions of 4×4, when the first prediction region is a prediction region located on the upper right side in the coding region, and the second prediction region is a prediction region located on the upper left side in the coding region, the first adjacent region is an adjacent region adjacent to the left side of the prediction region, and the second adjacent region is an adjacent region adjacent to the lower left side of the prediction region,
when the first prediction region is a prediction region located on the lower left side in the coding region, and the second prediction region is a prediction region located on the upper left side in the coding region, the first adjacent region is an adjacent region adjacent to the upper side of the prediction region, and the second adjacent region is an adjacent region adjacent to the upper right side of the prediction region, and
when the first prediction region is a prediction region located on the lower right side in the coding region, and the second prediction region is a prediction region located on the upper left side in the coding region, the first adjacent region is an adjacent region adjacent to the upper left side of the prediction region.
7. The image processing apparatus according to claim 2 ,
wherein the coding region is split into a plurality of prediction regions on the basis of asymmetric motion partition.
8. The image processing apparatus according to claim 2 , further comprising:
a reception unit that receives a coded stream and a flag indicating whether or not setting of the spatially adjacent region will be performed; and
a decoding unit that decodes the coded stream received by the reception unit so as to generate the image,
wherein the adjacent region setting unit performs setting of the spatially adjacent region on the basis of the flag received by the reception unit, and
wherein the reception unit and the decoding unit are each implemented via at least one processor.
9. The image processing apparatus according to claim 8 ,
wherein the flag is set for each coding region or the prediction region.
10. The image processing apparatus according to claim 2 ,
wherein whether or not setting of the spatially adjacent region will be performed is set on the basis of a sequence profile or level, and
wherein the adjacent region setting unit performs setting of the spatially adjacent region on the basis of the sequence profile or level.
11. An image processing method comprising:
causing an image processing apparatus to
set spatially adjacent regions on the basis of positions of prediction regions in a coding region so that generation of spatial motion vectors of the plurality of prediction regions is performed in parallel when a spatial prediction motion vector is generated among prediction motion vectors used to decode motion vectors of the plurality of prediction regions forming the coding region of an image;
generate a spatial prediction vector of the prediction region by using the motion vector of the set spatially adjacent region; and
decode a motion vector of the prediction region by using the prediction motion vector of the prediction region,
wherein, in a case where a first prediction region is located on a right or lower side in the coding region, a first adjacent region of a second prediction region is set instead of the first adjacent region serving as the second prediction region located on a left or upper side in the coding region among adjacent regions of the first prediction region.
12. An image processing apparatus comprising:
an adjacent region setting unit that sets spatially adjacent regions on the basis of positions of prediction regions in a coding region so that generation of spatial motion vectors of the plurality of prediction regions is performed in parallel when a spatial prediction motion vector is generated among prediction motion vectors used to code motion vectors of the plurality of prediction regions forming the coding region of an image;
a prediction motion vector generation unit that generates a spatial prediction vector of the prediction region by using the motion vector of the spatially adjacent region set by the adjacent region setting unit; and
a motion vector coding unit that codes a motion vector of the prediction region by using the prediction motion vector of the prediction region,
wherein, in a case where a first prediction region is located on a right or lower side in the coding region, the adjacent region setting unit sets a first adjacent region of a second prediction region instead of the first adjacent region serving as the second prediction region located on a left or upper side in the coding region among adjacent regions of the first prediction region, and
wherein the adjacent region setting unit, the prediction motion vector generation unit, and the motion vector coding unit are each implemented via at least one processor.
13. The image processing apparatus according to claim 12 ,
wherein, in a case where the first prediction region is located on the right or lower side in the coding region, the adjacent region setting unit sets a second adjacent region of the second prediction region instead of the second adjacent region adjacent to the first prediction region among the adjacent regions of the first prediction region.
14. The image processing apparatus according to claim 13 ,
wherein, in a case where the coding region is split into prediction regions of 2N×N, the first prediction region is a prediction region located on the lower side in the coding region, the second prediction region is a prediction region located on the upper side in the coding region, the first adjacent region is an adjacent region adjacent to an upper side of the prediction region, and the second adjacent region is an adjacent region adjacent to an upper right side of the prediction region.
15. The image processing apparatus according to claim 13 ,
wherein, in a case where the coding region is split into prediction regions of N×2N, the first prediction region is a prediction region located on the right side in the coding region, the second prediction region is a prediction region located on the left side in the coding region, the first adjacent region is an adjacent region adjacent to a left side of the prediction region, and the second adjacent region is an adjacent region adjacent to a lower left side of the prediction region.
16. The image processing apparatus according to claim 13 ,
wherein, in a case where the first prediction region is located on the lower right side in the coding region, the adjacent region setting unit sets a first adjacent region of a second prediction region instead of the first adjacent region serving as the second prediction region located on the upper left side in the coding region among adjacent regions of the first prediction region.
17. The image processing apparatus according to claim 16 ,
wherein, in a case where the coding region is split into prediction regions of 4×4, when the first prediction region is a prediction region located on the upper right side in the coding region, and the second prediction region is a prediction region located on the upper left side in the coding region, the first adjacent region is an adjacent region adjacent to the left side of the prediction region, and the second adjacent region is an adjacent region adjacent to the lower left side of the prediction region,
when the first prediction region is a prediction region located on the lower left side in the coding region, and the second prediction region is a prediction region located on the upper left side in the coding region, the first adjacent region is an adjacent region adjacent to the upper side of the prediction region, and the second adjacent region is an adjacent region adjacent to the upper right side of the prediction region, and
when the first prediction region is a prediction region located on the lower right side in the coding region, and the second prediction region is a prediction region located on the upper left side in the coding region, the first adjacent region is an adjacent region adjacent to the upper left side of the prediction region.
18. The image processing apparatus according to claim 13 ,
wherein the coding region is split into a plurality of prediction regions on the basis of asymmetric motion partition.
19. The image processing apparatus according to claim 13 , further comprising:
a setting unit that sets a flag indicating whether or not setting of the spatially adjacent region will be performed;
a coding unit that codes the image so as to generate a coded stream; and
a transmission unit that transmits the motion vector coded by the motion vector coding unit, the coded stream generated by the coding unit, and the flag set by the setting unit,
wherein the adjacent region setting unit performs setting of the spatially adjacent region on the basis of the flag set by the setting unit, and
wherein the setting unit, the coding unit, and the transmission unit are each implemented via at least one processor.
20. The image processing apparatus according to claim 19 ,
wherein the setting unit sets the flag for each coding region or the prediction region.
21. The image processing apparatus according to claim 13 ,
wherein whether or not setting of the spatially adjacent region will be performed is set on the basis of a sequence profile or level, and
wherein the adjacent region setting unit performs setting of the spatially adjacent region on the basis of the sequence profile or level.
22. An image processing method comprising:
causing an image processing apparatus to
set spatially adjacent regions on the basis of positions of prediction regions in a coding region so that generation of spatial motion vectors of the plurality of prediction regions is performed in parallel when a spatial prediction motion vector is generated among prediction motion vectors used to code motion vectors of the plurality of prediction regions forming the coding region of an image;
generate a spatial prediction vector of the prediction region by using the motion vector of the set spatially adjacent region; and
code a motion vector of the target region by using the prediction motion vector of the prediction region,
wherein, in a case where a first prediction region is located on a right or lower side in the coding region, a first adjacent region of a second prediction region is set instead of the first adjacent region serving as the second prediction region located on a left or upper side in the coding region among adjacent regions of the first prediction region.