Method for encoding and decoding images, and device using same

1. A method for inter prediction, comprising:

receiving mode information on an inter prediction of a current block, wherein the mode information includes merge flag information indicating whether an inter prediction mode for the current block is a merge mode and merge index information indicating a block used for deriving motion information of the current block;

deriving candidate blocks for the merge mode of the current block, wherein the candidate blocks comprises neighboring blocks of the current block;

deriving motion information of the current block by setting motion vector of the current block equal to motion vector of the block indicated by the merge index among the candidate blocks; and

performing inter prediction on the current block by using the motion vector of the current block,

wherein the neighboring blocks includes at least one of a left neighboring block of the current block, a top neighboring block of the current block, a right top corner neighboring block of the current block, a left top corner neighboring block of the current block, a left bottom corner neighboring block of the current block.

2. The method of claim 1 , wherein the mode information includes residual flag information indicating whether there is a residual signal for the current block, and

wherein the residual flag information is checked prior to the merge flag information.

3. The method of claim 2 , wherein the candidate blocks has the same configuration regardless whether there is the residual signal for the current block.

4. The method of claim 2 , wherein the derived motion information is one of L0 motion information for forward prediction, L1 motion information for backward prediction, and Bi motion information for bi prediction.

5. The method of claim 2 , wherein the left neighboring block is a bottom-most block among left neighboring blocks adjacent to the left boundary of the current block, and the top neighboring block is a right-most block among top neighboring blocks adjacent to the top boundary of the current block.

6. The method of claim 1 , wherein the candidate blocks comprises a co-located block which is a block at the same position as the current block among blocks in a reference picture.

7. The method of claim 1 , wherein decoding the received mode information comprises:

decoding the merge flag information prior to the residual flag information; and

decoding the residual flag information only when the decoded merge flag information indicates that the prediction mode for the current block is a merge mode.