Method for encoding and decoding coding unit
Initial value is selected when motion estimation is performed using predicted motion vector in video codec in which multiple references are allowed, and coding amount of motion vectors is reduced using predicted motion vector. Motion vector encoding apparatus includes predicted motion vector generator that generates adjacent motion vectors of current block with respect to each of reference pictures and generates each predicted motion vector of the current block from the generated adjacent motion vectors, motion estimator that determines final motion vector of the current block and final reference picture corresponding to the final motion vector with regard to search regions of the reference pictures, differential unit that obtains difference between the final motion vector of the current block and final predicted motion vector corresponding to the determined final reference picture among the predicted motion vectors, and unit for encoding information about the final reference picture and the difference.
1. An encoding method of an encoding device, comprising:
dividing a picture into a plurality of coding units of a recursive tree structure;
partitioning a leaf coding unit of the recursive tree structure into a plurality of prediction units;
determining a motion vector for a first prediction unit among the plurality of prediction units;
generating a prediction motion vector for the first prediction unit based on two or more motion vectors of neighboring prediction units adjacent to the first prediction unit, wherein the neighboring prediction units refer to a same reference picture;
generating a residual signal based on the first prediction unit and a prediction signal indicated by the motion vector for the first prediction unit;
encoding the residual signal; and
encoding a difference between the motion vector for the first prediction unit and the prediction motion vector for the first prediction unit,
wherein the leaf coding unit is indivisible into smaller coding units,
wherein both a maximum allowable hierarchical depth of the recursive tree structure and a size of a smallest coding unit for the recursive tree structure are embedded in a sequence parameter set, and
wherein a minimum size of allowable sizes for the smallest coding unit for the recursive tree structure is 8.
2. The encoding method of claim 1 , wherein the recursive tree structure is represented by a series of flags.
3. The encoding method of claim 2 , wherein the flag represents whether or not a coding unit with depth of k is divided into four coding units with depth of k+1.
4. The encoding method of claim 3 , wherein the flag set to 1 represents that a coding unit with depth of k is divided into four coding units with depth of k+1, and
the flag set to 0 represents that a coding unit with depth of k is a leaf coding unit indivisible into smaller coding units.
5. The encoding method of claim 1 , wherein the leaf coding unit is symmetrically partitioned into the plurality of prediction units for intra prediction, and
the leaf coding unit is asymmetrically partitioned into the plurality of prediction units for inter prediction.
6. A decoding method of a decoding device, comprising:
obtaining a leaf coding unit from a recursive tree structure including a plurality of coding units, wherein the leaf coding unit is indivisible into smaller coding units;
obtaining a prediction unit from the leaf coding unit;
obtaining a motion vector difference for the prediction unit from an input bitstream;
generating a prediction motion vector for the prediction unit based on two or more motion vectors of neighboring prediction units adjacent to the prediction unit, wherein the neighboring prediction units refer to a same reference frame;
adding the motion vector difference and the prediction motion vector to generate a motion vector for the prediction unit; and
generating a residual signal between the prediction unit and a prediction signal indicated by the motion vector for the prediction unit by decoding the input bitstream;
wherein both a maximum allowable hierarchical depth of the recursive tree structure and a size of a smallest coding unit for the recursive tree structure are signaled from a sequence parameter set, and
wherein a minimum size of allowable sizes for the smallest coding unit for the recursive tree structure is 8.
7. The decoding method of claim 6 , wherein the recursive tree structure is represented by a series of flags.
8. The decoding method of claim 7 , wherein the flag represents whether or not a coding unit with depth of k is divided into four coding units with depth of k+1.
9. The decoding method of claim 8 , wherein the flag set to 1 represents that a coding unit with depth of k is divided into four coding units with depth of k+1, and
the flag set to 0 represents that a coding unit with depth of k is a leaf coding unit indivisible into smaller coding units.
10. The decoding method of claim 6 , wherein the obtained prediction unit is a symmetrically partitioned unit when the obtained prediction unit is intra prediction unit, and
the obtained prediction unit is an asymmetrically partitioned unit when the obtained prediction unit is inter prediction unit.