Method and apparatus for motion vector encoding/decoding using spatial division, and method and apparatus for image encoding/decoding using same
Disclosed is a method for encoding a motion vector by using spatial division, the method comprises: aggregating predicted motion vector candidates of a current block into a plurality of groups; selecting representative predicted motion vectors for respective groups by selecting one of aggregated predicted motion vector candidates within the respective groups as representative predicted motion vector; selecting single representative predicted motion vector from the representative predicted motion vectors for the respective groups as a predicted motion vector; and encoding a differential vector representing difference between a current motion vector of the current block and a selected predicted motion vector.
1. An apparatus for encoding a current motion vector of a current block in a current frame, the apparatus comprising:
a motion vector selector configured to determine a plurality of motion vector candidates derived from first multiple neighboring positions in the current frame and second multiple neighboring positions in the current frame by
deriving a first motion vector candidate of the plurality of motion vector candidates by using first multiple neighboring motion vectors corresponding respectively to the first multiple neighboring positions, and
deriving a second motion vector candidate of the plurality of motion vector candidates by using second multiple neighboring motion vectors corresponding respectively to the second multiple neighboring positions,
wherein the first multiple neighboring positions and the second multiple neighboring positions are adjacent to the current block;
a predicted motion vector selector configured to
select a motion vector among the plurality of motion vector candidates, and
set the motion vector selected among the plurality of motion vector candidates, as a predicted motion vector of the current motion vector;
a differential vector encoder configured to encoded a differential vector representing a difference between the current motion vector and the predicted motion vector, and
a predicted motion vector encoder configured to encode an index for identifying the motion vector selected among the plurality of motion vector candidates,
wherein each of the first multiple neighboring positions is different from each of the second multiple neighboring positions.
2. An apparatus for decoding a current motion vector of a current block in a current frame, the apparatus comprising:
a motion vector selector configured to determine a plurality of motion vector candidates derived from first multiple neighboring positions in the current frame and second multiple neighboring positions in the current frame by
deriving a first motion vector candidate of the plurality of motion vector candidates by using first multiple neighboring motion vectors corresponding respectively to the first multiple neighboring positions, and
deriving a second motion vector candidate of the plurality of motion vector candidates by using second multiple neighboring motion vectors corresponding respectively to the second multiple neighboring positions,
wherein the first multiple neighboring positions and the second multiple neighboring positions are adjacent to the current block;
a predicted motion vector reconstructor configured to
reconstruct an index for reconstructing a predicted motion vector of the current motion vector by decoding an index data extracted from a bitstream
select a motion vector among the plurality of motion vector candidates, and
set the motion vector selected among the plurality of motion vector candidates, as a predicted motion vector of the current motion vector; and
a current motion vector reconstructor configured to
reconstruct a differential vector representing a difference between the current motion vector and the predicted motion vector by decoding a differential vector data extracted from the bitstream, and
reconstruct the current motion vector of the current block by adding the reconstructed differential vector and the predicted motion vector,
wherein each of the first multiple neighboring positions is different from each of the second multiple neighboring positions.
3. A method of encoding a current motion vector of a current block in a current frame, the method comprising:
determining a plurality of motion vector candidates derived from first multiple neighboring positions in the current frame and second multiple neighboring positions in the current frame by
deriving a first motion vector candidate of the plurality of motion vector candidates by using first multiple neighboring motion vectors corresponding respectively to the first multiple neighboring positions, and
deriving a second motion vector candidate of the plurality of motion vector candidates by using second multiple neighboring motion vectors corresponding respectively to the second multiple neighboring positions,
wherein the first multiple neighboring positions and the second multiple neighboring positions are adjacent to the current block;
selecting a motion vector among the plurality of motion vector candidates;
setting the motion vector selected among the plurality of motion vector candidates, as a predicted motion vector of the current motion vector;
encoding a differential vector representing a difference between a current motion vector and the predicted motion vector; and
encoding an index for identifying the motion vector selected among the plurality of motion vector candidates,
wherein each of the first multiple neighboring positions is different from each of the second multiple neighboring positions.
4. The method of claim 3 , wherein the deriving of the first motion vector comprises:
determining a motion vector from the first multiple neighboring motion vectors, based on a spatial distance from a median of the first multiple neighboring motion vectors to the motion vector determined from the first multiple neighboring motion vectors, and
setting the motion vector, determined from the first multiple neighboring motion vectors, as the first motion vector candidate.
5. The method of claim 3 , wherein the predicted motion vector is generated from the plurality of motion vectors by using rate-distortion costs.
6. The method of claim 3 , wherein a median a plurality of neighboring motion vectors including the first multiple neighboring motion vectors and the second multiple neighboring motion vectors is selected as the predicted motion vector, when a spatial distance between the plurality of groups of neighboring motion vectors is equal to or shorter than a predetermined reference distance.
7. A method of decoding a current motion vector of a current block in a current frame, the method comprising:
reconstructing a differential vector representing a difference between the current motion vector and a predicted motion vector of the current motion vector by decoding a differential vector data from a bitstream;
reconstructing an index for reconstructing the predicted motion vector by decoding an index data extracted from the bitstream;
determining a plurality of motion vector candidates derived from first multiple neighboring positions in the current frame and second multiple neighboring positions in the current frame, both of which are adjacent to the current block, by
deriving a first motion vector candidate of the plurality of motion vector candidates by using first multiple neighboring motion vectors corresponding respectively to the first multiple neighboring positions, and
deriving a second motion vector candidate of the plurality of motion vector candidates by using second multiple neighboring motion vectors corresponding respectively to the second multiple neighboring positions,
wherein the first multiple neighboring positions and the second multiple neighboring positions are adjacent to the current block;
selecting a motion vector identified by the index among the plurality of motion vector candidates;
setting the motion vector selected among the plurality of motion vector candidates, as the predicted motion vector of the current motion vector; and
reconstructing the current motion vector of the current block by adding the reconstructed differential vector and the predicted motion vector;
wherein each of the first multiple neighboring positions is different from each of the second multiple neighboring positions.
8. The method of claim 7 , wherein the deriving of the first motion vector comprises:
determining a motion vector from the first multiple neighboring motion vectors, based on a spatial distance from a median of the first multiple neighboring motion vectors to the motion vector determined from the first multiple neighboring motion vectors, and
setting the motion vector, determined from the first multiple neighboring motion vectors, as the first motion vector candidate.
9. The method of claim 7 , further comprising:
determining whether the index is extracted from the bitstream; and
selecting a median of the plurality of neighboring predicted motion as the predicted motion vector, when the index is not extracted from the bitstream.
10. The apparatus of claim 2 , wherein one of the second multiple neighboring positions is in a reference frame reconstructed prior to the current frame, and said one of the second multiple neighboring positions in the reference frame is determined based on a position of the current block in the current frame.