Method for encoding/decoding image signal, and device therefor

A method of encoding a video, the method comprising:

• determining a partitioning type of a coding block;
• partitioning the coding block into a first prediction unit and a second prediction unit based on the determined partitioning type,
• wherein the partitioning type is determined as one of a plurality of partitioning type candidates, and
• wherein the plurality of partitioning type candidates comprises a symmetric partitioning type that a size of the first prediction unit and a size of the second prediction unit are the same and an asymmetric partitioning type that the size of the first prediction unit is different from the size of the second prediction unit;
• deriving motion information of each of the first prediction unit and the second prediction unit;
• and
• obtaining a prediction sample for the first prediction unit and the second prediction unit based on the derived motion information,
• wherein a first merge index for specifying a first merge candidate used to derive the motion information of the first prediction unit and a second merge index for specifying a second merge candidate used to derive the motion information of the second prediction unit is encoded, respectively.

A device for decoding a video, the device comprising:

• an inter prediction unit configured to:
• determine a partitioning type of a coding block, partition the coding block into a first prediction unit and a second prediction unit based on the determined partitioning type, derive motion information of each of the first prediction unit and the second prediction unit, and obtain a prediction sample for the first prediction unit and the second prediction unit based on the derived motion information,
• wherein the partitioning type is determined by index information specifying one of a plurality of partitioning type candidates, and
• wherein the plurality of partitioning type candidates comprises a symmetric partitioning type that a size of the first prediction unit and a size of the second prediction unit are the same and an asymmetric partitioning type that the size of the first prediction unit is different from the size of the second prediction unit, and
• wherein the motion information of the first prediction unit is derived from a first merge candidate specified by a first merge index signaled in a bitstream and the motion information of the second prediction unit is derived from a second merge candidate specified by a second merge index signaled in the bitstream.