Back to results
US 12,627,841 B2Granted
Video encoding method, video decoding method, non-transitory computer-readable medium
AssigneeGuangdong Oppo Mobile Telecommunications Corp., Ltd.
InventorsYue Yu, Haoping Yu
Life of this patent
Estimating the term…
Abstract
A video encoding method includes: dividing, based on a size of region, a plurality of quantization levels comprising a last non-zero quantization level within a block of a video into a plurality of regions in a level coding scan order to select a plurality of defined regions among the plurality of regions based on the last non-zero quantization level; generating, based on the quantization levels within the plurality of defined regions in a predefined scan order, a plurality of syntax structures associated with the plurality of defined regions; and encoding the plurality of syntax structures associated with the plurality of defined regions into a bitstream.
Claims18
A video encoding method, comprising:
- dividing, based on a size of region, a plurality of quantization levels comprising a last non-zero quantization level within a block of a video into a plurality of regions in a level coding scan order to select a plurality of defined regions among the plurality of regions based on the last non-zero quantization level;
- generating, based on the quantization levels within the plurality of defined regions in a predefined scan order, a plurality of syntax structures associated with the plurality of defined regions;
- and
- encoding the plurality of syntax structures associated with the plurality of defined regions into a bitstream;
- wherein the generating, based on the quantization levels within the plurality of defined regions in the predefined scan order, the plurality of syntax structures associated with the plurality of defined regions comprises:
- in response to determining that the predefined scan order is a forward scan order from a first position index within the block to a position index of the last non-zero quantization level within the block, generating the plurality of syntax structures based on the quantization levels within the plurality of defined regions from a first-forward defined region associated with the first position index to a last-forward defined region associated with a position index of the last non-zero quantization level, region by region;
- and
- in response to determining that the predefined scan order is a reverse scan order from the position index of the last non-zero quantization level within the block to the first position index within the block, generating the plurality of syntax structures based on the quantization levels within the plurality of defined regions from a first-reverse defined region associated with the position index of the last non-zero quantization level to a last-reverse defined region associated with the first position index, region by region.
5 more elements
A video decoding method, comprising:
- receiving a plurality of syntax structures associated with a plurality of defined regions from a bitstream of a video;
- recovering, based on the plurality of syntax structures and a size of region, a plurality of quantization levels comprising a last non-zero quantization level within the plurality of defined regions in a predefined scan order;
- and
- reconstructing the plurality of quantization levels within the plurality of defined regions among a plurality of regions within a block of the video in a level coding scan order;
- wherein the recovering, based on the plurality of syntax structures and the size of region, the plurality of quantization levels comprising the last non-zero quantization level within the plurality of defined regions in the predefined scan order comprises:
- in response to determining that the predefined scan order is a forward scan order from a first position index within the block to a position index of the last non-zero quantization level within the block, recovering the quantization levels within the plurality of defined regions from a first-forward defined region associated with the first position index to a last-forward defined region associated with a position index of the last non-zero quantization level, region by region;
- and
- in response to determining that the predefined scan order is a reverse scan order from the position index of the last non-zero quantization level within the block to the first position index within the block, recovering the quantization levels within the plurality of defined regions from a first-reverse defined region associated with the position index of the last non-zero quantization level within the block to a last-reverse defined region associated with the first position index, region by region.
5 more elements
Dependencies are inferred from the claim text.