Method for encoding and decoding in parallel processing and apparatus using same
AssigneeLG Electronics Inc.
InventorsHendry Hendry, Byeongmoon Jeon, Jungsun Kim, Sangoh Jeong
Life of this patent
Estimating the term…
Abstract
A video encoding method includes a step of encoding substreams which are rows of largest coding units (LCUs) in parallel with each other, and a step of transmitting a bit stream including the encoded substreams, where the number of the substreams may be the same as the number of entry points.
Claims14
A video decoding method, by a video decoding apparatus, comprising:
- receiving a bitstream including information on a slice header and information on substreams for a current slice segment;
- obtaining entry point information for the substreams from the slice header;
- and
- decoding the substreams based on the entry point information to reconstruct a picture, wherein decoding the substreams includes:
- deriving prediction information and residual information on a block of a current substream;
- deriving prediction samples on the block based on the prediction information;
- deriving residual samples on the block based on the residual information;
- generating reconstructed samples to generate the reconstructed picture based on the prediction samples and the residual samples;
- and
- performing deblocking filtering on the reconstructed picture,
- wherein the picture includes multiple largest coding units (LCUs),
- wherein a number of the substreams is equal to a number of LCU rows in the current slice segment in the picture,
- wherein the entry point information includes number information indicating a number of entry point offsets, and
- wherein the number of the substreams is derived based on the number information in the slice header.
A video encoding method, by a video encoding apparatus, comprising:
- encoding substreams for a current slice segment in a picture;
- and
- generating a bitstream including information on a slice header and information on substreams for the current slice segment,
- wherein the slice header includes entry point information for the substreams,
- wherein the picture includes multiple largest coding units (LCUs),
- wherein a number of the substreams is equal to a number of LCU rows in the current slice segment in the picture,
- wherein the entry point information includes number information indicating a number of entry point offsets, and
- wherein the number of the substreams is derived based on the number information in the slice header.
A non-transitory decoder-readable storage medium storing a bitstream comprising a decoder executable program, the decoder executable program, when executed, causing a decoder to perform the following steps:
- obtaining entry point information for substreams from a slice header, wherein the bitstream includes information on the slice header and information on the substreams for a current slice segment;
- and
- decoding the substreams based on the entry point information to reconstruct a picture,
- wherein decoding the substreams includes:
- deriving prediction information and residual information on a block of a current substream;
- deriving prediction samples on the block based on the prediction information;
- deriving residual samples on the block, wherein the residual samples are derived based on the residual information;
- generating reconstructed samples to generate the reconstructed picture based on the prediction samples and the residual samples;
- and
- performing deblocking filtering on the reconstructed picture,
- wherein the picture includes multiple largest coding units (LCUs),
- wherein a number of the substreams is equal to a number of LCU rows in the current slice segment in the picture,
- wherein the entry point information includes number information indicating a number of entry point offsets, and
- wherein the number of the substreams is derived based on the number information in the slice header.
Dependencies are inferred from the claim text.