Method and apparatus for parallel video decoding based on multi-core system
AssigneeSamsung Electronics Co., Ltd.
InventorsHyung-ju Chun, Chan-sik Park, Ki-won Yoo, Jae-moon Kim
Life of this patent
Estimating the term…
Abstract
A video decoding method includes obtaining a motion vector of a current block belonging to a first picture from a bitstream, performed by a first decoding unit; determining whether a reference block indicated by the motion vector is decoded, performed by the first decoding unit; and decoding the current block, based on whether the reference block is decoded. The reference block is included in a second picture decoded by a second decoding unit. The first picture and the second picture are decoded in parallel.
Claims8
A video decoding method comprising:
- obtaining a motion vector of a current block belonging to a first picture from a bitstream, performed by a first decoding unit;
- determining whether a reference block indicated by the motion vector is decoded, performed by the first decoding unit;
- and
- when the reference block is not decoded, modifying the motion vector, performed by the first decoding unit;
- and
- decoding the current block by using the modified motion vector, performed by the first decoding unit,
- wherein the reference block is included in a second picture decoded by a second decoding unit, and
- the first picture and the second picture are decoded in parallel.
A video decoding apparatus for performing parallel-decoding, the video decoding apparatus comprising:
- a first decoding unit configured to obtain a motion vector of a current block belonging to a first picture from a bitstream, determine whether a reference block indicated by the motion vector is decoded, modifying the motion vector when the reference block is not decoded, and decode the current block based on whether the reference block is decoded by using the modified motion vector;
- and a second decoding unit configured to decode a second picture including the reference block, and the first picture and the second picture are decoded in parallel.
Dependencies are inferred from the claim text.