Method for processing image on basis of inter-prediction mode and apparatus therefor
Disclosed are a method for decoding a video signal and an apparatus therefor. Specifically, a method for decoding an image based on an inter-prediction mode may include: configuring a merge candidate list by using motion information of a spatial candidate and a temporal candidate adjacent to a current block; parsing a first syntax element indicating a particular candidate applied to inter-prediction of the current block among a predefined number of candidates in the merge candidate list; deriving a refined candidate by adding a motion vector offset to a motion vector of a candidate indicated by the first syntax element; and generating a prediction block of the current block by using the motion information of the refined candidate.
1 . A decoding apparatus for image decoding, the decoding apparatus comprising:
a memory; and
at least one processor connected to the memory, the at least one processor configured to:
obtain residual information of a current block from a bitstream;
determine an inter prediction mode of the current block as a merge mode among the merge mode and a Motion Vector Prediction (MVP) mode;
configure a merge candidate list based on motion information of a spatial candidate and a temporal candidate for the current block in the merge mode;
obtain merge motion vector offset precision information related to whether a merge motion vector offset has a fractional pixel precision, wherein the merge motion vector offset precision information is obtained from a picture parameter set;
determine a merge candidate used to derive a refined merge candidate based on the merge motion vector offset, wherein a merge candidate on 0th position of the merge candidate list is determined as the merge candidate used to derive a refined merge candidate based on the merge motion vector offset;
derive motion information of the refined merge candidate based on a sum of the merge motion vector offset, which is derived based on the merge motion vector offset precision information, and a motion vector of the determined merge candidate; and
derive a prediction block of the current block based on the motion information of the refined merge candidate; and
derive a residual block of the current block based on the residual information; and
generate a reconstructed block of the current block based on the prediction block and the residual block,
wherein a merge candidate generated based on an average of a pair of predefined candidates is added in the merge candidate list after configuring the merge candidate list including the spatial candidate and the temporal candidate,
wherein the pair of predefined candidates is a pair of a candidate on the 0th position and a candidate on 1st position of the merge candidate list, and
wherein the merge motion vector offset is derived based on a value acquired by multiplying a merge motion vector offset precision by a non-zero integer.
2 . An encoding apparatus for image encoding, the encoding apparatus comprising:
a memory; and
at least one processor connected to the memory, the at least one processor configured to:
determine an inter prediction mode of a current block as a merge mode among the merge mode and a Motion Vector Prediction (MVP) mode;
configure a merge candidate list based on motion information of a spatial candidate and a temporal candidate for the current block in the merge mode;
determine a merge candidate used to derive a refined merge candidate based on a merge motion vector offset, wherein a merge candidate on 0th position of the merge candidate list is determined as the merge candidate used to derive a refined merge candidate based on the merge motion vector offset;
derive motion information of the refined merge candidate based on a sum of the merge motion vector offset, which is derived based on merge motion vector offset precision information, and a motion vector of the determined merge candidate;
derive a prediction block of the current block based on the motion information of the refined merge candidate;
derive a residual block of the current block based on the prediction block of the current block; and
generate video information including the merge motion vector offset precision information related to whether the merge motion vector offset has a fractional pixel precision and residual information for the residual block, wherein the merge motion vector offset precision information is included in a picture parameter set of the video information,
wherein a merge candidate generated based on an average of a pair of predefined candidates is added in the merge candidate list after configuring the merge candidate list including the spatial candidate and the temporal candidate,
wherein the pair of predefined candidates is a pair of a candidate on the 0th position and a candidate on 1st position of the merge candidate list, and
wherein the merge motion vector offset is derived based on a value acquired by multiplying a merge motion vector offset precision by a non-zero integer.
3 . An apparatus for transmitting data for an image, the apparatus comprising:
at least one processor configured to obtain a bitstream of video information including merge motion vector offset precision information related to whether a merge motion vector offset has a fractional pixel precision and residual information for a residual block of the current block, wherein the merge motion vector offset precision information is included in a picture parameter set of the video information; and
a transmitter configured to transmit the data including the bitstream of the video information including the merge motion vector offset precision information and the residual information,
wherein a prediction block of the current block is derived by determining an inter prediction mode of the current block as a merge mode among the merge mode and a Motion Vector Prediction (MVP) mode, configuring a merge candidate list based on motion information of a spatial candidate and a temporal candidate for the current block in the merge mode, determining a merge candidate used to derive a refined merge candidate based on the merge motion vector offset, wherein a merge candidate on 0th position of the merge candidate list is determined as the merge candidate used to derive a refined merge candidate based on the merge motion vector offset, deriving motion information of the refined merge candidate based on a sum of the merge motion vector offset, which is derived based on the merge motion vector offset precision information, and a motion vector of the determined merge candidate,
wherein a merge candidate generated based on an average of a pair of predefined candidates is added in the merge candidate list after configuring the merge candidate list including the spatial candidate and the temporal candidate,
wherein the pair of predefined candidates is a pair of a candidate on the 0th position and a candidate on 1st position of the merge candidate list, and
wherein the merge motion vector offset is derived based on a value acquired by multiplying a merge motion vector offset precision by a non-zero integer.