Image encoding/decoding method and apparatus, and recording media storing bitstream
An image encoding/decoding method is provided. An image decoding method of the present invention may comprise constructing a motion information list for a current block, reconstructing a first index for a first triangular block within the current block and a second index for a second triangular block within the current block, and predicting the first triangular block and the second triangular block, based on the first index, the second index, and the motion information list.
1. A method of decoding an image performed by an image decoding apparatus, the method comprising:
reconstructing a first index for a first non-rectangular block within the current block and a second index for a second non-rectangular block within the current block, each of the first index and the second index being related to selection to a motion information candidate from among multiple motion information candidates; and
predicting the first non-rectangular block and the second non-rectangular block, based on the first index and the second index,
wherein motion information on the first non-rectangular block or the second non-rectangular block is determined to be L0 motion information which a motion information candidate corresponding to the first index or the second index has, in response to the first index or the second index being an even number index, and
wherein motion information on the first non-rectangular block or the second non-rectangular block is determined to be L1 motion information which the motion information candidate corresponding to the first index or the second index has, in response to the first index or the second index being an odd number index.
2. The method of claim 1 , wherein the current block is a quadrangle, and
the first non-rectangular block and the second non-rectangular block are regions that are generated by partitioning the current block with a boundary line selected from a plurality of predefined boundary lines.
3. The method of claim 2 , wherein information on the selected boundary line is decoded from a bitstream.
4. The method of claim 1 , wherein the multiple motion information candidates comprises a predetermined number of pieces of motion information in an order of motion information of a spatially neighboring block, motion information of a temporally neighboring block, buffer-based motion information, combination motion information, and zero motion information.
5. The method of claim 1 , wherein the first index and the second index are different from each other.
6. The method of claim 1 , wherein when the motion information candidate corresponding to the first index or the second index does not include the determined motion information, a prediction direction of the motion information on the first non-rectangular block or the second non-rectangular block is changed into the opposite direction.
7. A method of encoding an image performed by an image encoding apparatus, the method comprising:
determining motion information for a first non-rectangular block within a current block and motion information for a second non-rectangular block within the current block;
predicting the first non-rectangular block and the second non-rectangular block, based on the motion information for the first non-rectangular block and the motion information for the second non-rectangular block; and
encoding a first index for the first non-rectangular block and a second index for the second non-rectangular block, based on the motion information for the first non-rectangular block and the motion information for the second non-rectangular block, each of the first index and the second index being related to selection to a motion information candidate from among multiple motion information candidates,
wherein motion information on the first non-rectangular block or the second non-rectangular block is determined to be L0 motion information which a motion information candidate corresponding to the first index or the second index has, in response to the first index or the second index being an even number index, and
wherein motion information on the first non-rectangular block or the second non-rectangular block is determined to be L1 motion information which the motion information candidate corresponding to the first index or the second index has, in response to the first index or the second index being an odd number index.
8. The method of claim 7 , wherein the current block is a quadrangle, and
the first non-rectangular block and the second non-rectangular block are regions that are generated by partitioning the current block with a boundary line selected from a plurality of predefined boundary lines.
9. The method of claim 8 , wherein information on the selected boundary line is encoded into a bitstream.
10. The method of claim 7 , wherein the multiple motion information comprises a predetermined number of pieces of motion information in an order of motion information of a spatially neighboring block, motion information of a temporally neighboring block, buffer-based motion information, combination motion information, and zero motion information.
11. The method of claim 7 , wherein the first index and the second index are different from each other.
12. The method of claim 7 , wherein when the motion information candidate corresponding to the first index or the second index does not include the determined motion information, a prediction direction of the motion information on the first non-rectangular block or the second non-rectangular block is changed into the opposite direction.
13. A non-transitory computer-readable medium storing instructions, which when executed by at least one processor perform a method for transmitting a bitstream including encoded video data, the method comprising:
generating the bitstream by encoding a current block in an image; and
transmitting the bitstream,
wherein generating the bitstream comprises:
determining motion information for a first non-rectangular block within a current block and motion information for a second non-rectangular block within the current block;
predicting the first non-rectangular block and the second non-rectangular block, based on the motion information for the first non-rectangular block and the motion information for the second non-rectangular block; and
encoding a first index for the first non-rectangular block and a second index for the second non-rectangular block, based on the motion information for the first non-rectangular block and the motion information for the second non-rectangular block, each of the first index and the second index being related to selection to a motion information candidate from among multiple motion information candidates,
wherein based on whether the first index or the second index is a predetermined index, a prediction direction of the first non-rectangular block or the second non-rectangular block is determined, and
wherein motion information on the first non-rectangular block or the second non-rectangular block is determined to be L0 motion information which a motion information candidate corresponding to the first index or the second index has, in response to the first index or the second index being an even number index, and
wherein motion information on the first non-rectangular block or the second non-rectangular block is determined to be L1 motion information which the motion information candidate corresponding to the first index or the second index has, in response to the first index or the second index being an odd number index.