Method and device for coding image/video on basis of prediction weighted table
A video decoding method by a video decoding apparatus includes: parsing, from a picture parameter set of a bitstream, a first flag related to whether a weighted prediction is applied to slices referring to the picture parameter set; parsing, from the picture parameter set, a second flag related to whether information about the weighted prediction is present in a picture header of the bitstream or is present in a slice header of the bitstream based on the first flag; parsing a prediction weighted table syntax from the picture header or the slice header based on the first flag and the second flag; generating prediction samples for the current block by performing weighted prediction for the current block in a current picture based on syntax elements in the prediction weighted table syntax; and generating reconstructed samples based on the prediction samples.
1. A decoding apparatus for an image decoding, the decoding apparatus comprising:
a memory; and
at least one processor connected to the memory, the at least one processor configured to:
parse a first flag related to whether weighted prediction is applied to P slices referring to a picture parameter set, from the picture parameter set of a bitstream;
parse a second flag related to whether the weighted prediction is applied to B slices referring to the picture parameter set, from the picture parameter set;
parse a third flag related to whether information on the weighted prediction is present in a picture header of the bitstream or in a slice header of the bitstream, from the picture parameter set, based on the first flag or the second flag;
parse a prediction weighted table syntax from the picture header or the slice header;
generate prediction samples for a current block by performing the weighted prediction on the current block in a current picture based on syntax elements in the prediction weighted table syntax; and
generate reconstructed samples based on the prediction samples,
wherein the prediction weighted table syntax is parsed from the picture header based on the third flag,
wherein whether L0 number information for the weighted prediction is parsed from the prediction weighted table syntax is determined based on the third flag,
wherein whether L1 number information for the weighted prediction is parsed from the prediction weighted table syntax is determined based on the third flag,
wherein the L0 number information represents a number of weighted reference pictures in a reference picture list 0, and
wherein the L1 number information represents a number of weighted reference pictures in a reference picture list 1.
2. An encoding apparatus for an image encoding, the encoding apparatus comprising:
a memory; and
at least one processor connected to the memory, the at least one processor configured to:
derive motion information on a current block;
perform weighted prediction on the current block based on the motion information; and
encode image information comprising a first flag, a second flag and information on the weighted prediction, wherein a third flag is comprised in the encoded image information based on the first flag or the second flag,
wherein the first flag is related to whether weighted prediction is applied to P slices referring to a picture parameter set, the second flag is related to whether the weighted prediction is applied to B slices referring to the picture parameter set, the third flag is related to whether information on the weighted prediction is present in a picture header, and the third flag is comprised in the picture parameter set based on the third flag being comprised in the encoded image information,
wherein the information on the weighted prediction comprises a prediction weighted table syntax,
wherein the prediction weighted table syntax is comprised in the picture header or the slice header based on the third flag,
wherein whether L0 number information for the weighted prediction is configured to be parsed from the prediction weighted table syntax is determined based on the third flag,
wherein whether L1 number information for the weighted prediction is configured to be parsed from the prediction weighted table syntax is determined based on the third flag,
wherein the L0 number information represents a number of weighted reference pictures in a reference picture list 0, and
wherein the L1 number information represents a number of weighted reference pictures in a reference picture list 1.