IP Verse

Methods And Apparatus For Picture Encoding And Decoding

Patent No. US11917146 (titled "Methods And Apparatus For Picture Encoding And Decoding") was filed by Thomson Licensing on Mar 13, 2018.

What is this patent about?

’146 is related to the field of video coding and decoding, specifically addressing the efficient handling of picture blocks located at or extending beyond the boundaries of a video frame. Modern video codecs divide pictures into blocks for processing, and the way these blocks are split (or not split) significantly impacts compression efficiency. Prior art solutions often restrict the splitting modes available for boundary blocks, leading to suboptimal coding performance.

The underlying idea behind ’146 is to improve coding efficiency at picture boundaries by selectively allowing various block splitting modes based on the block's position relative to the boundary. The core insight is that by checking whether split lines coincide with picture borders or ensuring that remaining block parts after splitting are multiples of a minimum block size, the encoder and decoder can intelligently enable or disable certain split modes. This allows for a more flexible and efficient representation of image data near the edges of the frame.

The claims of ’146 focus on a decoding method (and corresponding encoding method, apparatus, and storage medium) where, when a block extends beyond a picture boundary, the decoder decodes information to identify the current split mode from a set of split modes. The key is that this set is limited to modes where the split direction is parallel to the picture boundary . The information decoded includes a flag indicating the number of sub-blocks resulting from the split, enabling efficient signaling of the chosen split mode.

In practice, the invention involves determining if a current block extends beyond the picture boundary. If it does, the encoder evaluates available split modes (horizontal, vertical, quad-tree, etc.) and restricts the selection to those where the split lines are parallel to the boundary. The encoder then signals the chosen split mode using a flag indicating the number of resulting sub-blocks. The decoder performs the reverse process, using the received flag to determine the split mode and decode the block accordingly. This approach avoids inefficient splits that would result in sub-blocks extending outside the picture.

This method differentiates itself from prior approaches by allowing a more nuanced selection of split modes for boundary blocks. Instead of simply forcing a quad-tree split or disallowing certain splits altogether, ’146 intelligently enables or disables modes based on boundary alignment and minimum block size constraints. This leads to a better trade-off between coding efficiency and bitstream overhead, resulting in improved rate-distortion performance and potentially avoiding the need for padding or other workarounds to ensure picture dimensions are multiples of minimum block sizes. The use of a flag to signal the number of sub-blocks further optimizes the signaling overhead.

How does this patent fit in bigger picture?

Technical landscape at the time

In the late 2010s when ’146 was filed, video coding schemes commonly employed prediction and transform techniques to achieve high compression efficiency. At a time when video coding standards typically partitioned pictures into coding tree units (CTUs) and coding units (CUs) using quad-tree structures, hardware or software constraints made flexible block partitioning schemes non-trivial.

Novelty and Inventive Step

The examiner approved the application because the claims require splitting a block by applying a split mode from a set of split modes. Each split mode specifies a single partitioning of an initial block into multiple second blocks, where each second block has a width and height equal to a predefined subdivision of the initial block's width and height. The split mode of the current block is determined from the set of split modes, and the current block is coded according to the split mode. When the current block extends beyond a picture boundary, determining the split mode includes decoding information that identifies the split mode from the set that is parallel to the boundary. This information includes a flag indicating the number of blocks resulting from the split. The examiner stated that prior art references disclose splitting blocks that cross a picture boundary, but do not disclose performing a single partitioning using a flag that indicates the number of blocks resulting from the split.

Claims

This patent contains 20 claims, with independent claims numbered 1, 4, 7, 10, 13, 15, 17, and 19. The independent claims generally focus on methods, apparatuses, and media for decoding and encoding a current block by determining and applying a split mode, particularly when the block extends beyond a picture boundary. The dependent claims generally elaborate on specific split modes and conditions for applying them.

Key Claim Terms New

Definitions of key terms used in the patent claims.

Term (Source)Support for SpecificationInterpretation
Current split mode
(Claim 1, Claim 4, Claim 7, Claim 10, Claim 13, Claim 15, Claim 17, Claim 19)		“A decoding method for decoding a current block, the current block including at least a portion of a picture, is disclosed that comprises: determining whether the current block extends beyond a boundary of the picture; determining for each split mode of a plurality of split modes whether the split mode is allowed for the current block by checking whether at least one of the split lines is co-located with one of the picture boundaries or whether a size of a block part, i.e. a remaining part resulting from the splitting, inside the picture along the picture boundary is a multiple of a minimum block size; decoding from a bitstream a current split mode of the current block responsive to the allowed split modes; and decoding the current block according to the current split mode.”A mode selected from a set of split modes that specifies how a first block (current block) is decomposed into a plurality of second blocks.
Number of second blocks
(Claim 1, Claim 4, Claim 7, Claim 10, Claim 13, Claim 15, Claim 17, Claim 19)		“According to a specific embodiment, the plurality of split modes comprises at least two of the following split modes, herein h is the height of the current block and w is the width of the current block: splitting the block horizontally into one upper sub-block of height h/4 and one bottom sub-block of height 3h/4; splitting the block horizontally into one bottom sub-block of height h/4 and one upper sub-block of height 3h/4; splitting the block vertically into one left sub-block of width w/4 and one right sub-block of width 3w/4; splitting the block vertically into one right sub-block of width w/4 and one left sub-block of width 3w/4; splitting the block horizontally into one sub-block of height h/4, one central sub-block of height h/2 and one sub-block of height h/4; splitting the block vertically into one sub-block of width w/4, one central sub-block of width w/2 and one sub-block of width w/4; splitting the block horizontally into two sub-blocks of height h/2; and splitting the block vertically into two sub-blocks of width w/2.”The quantity of blocks that result from splitting the current block according to the current split mode.
Set of split modes
(Claim 1, Claim 4, Claim 7, Claim 10, Claim 13, Claim 15, Claim 17, Claim 19)		“According to a specific embodiment, the plurality of split modes comprises at least two of the following split modes, herein h is the height of the current block and w is the width of the current block: splitting the block horizontally into one upper sub-block of height h/4 and one bottom sub-block of height 3h/4; splitting the block horizontally into one bottom sub-block of height h/4 and one upper sub-block of height 3h/4; splitting the block vertically into one left sub-block of width w/4 and one right sub-block of width 3w/4; splitting the block vertically into one right sub-block of width w/4 and one left sub-block of width 3w/4; splitting the block horizontally into one sub-block of height h/4, one central sub-block of height h/2 and one sub-block of height h/4; splitting the block vertically into one sub-block of width w/4, one central sub-block of width w/2 and one sub-block of width w/4; splitting the block horizontally into two sub-blocks of height h/2; and splitting the block vertically into two sub-blocks of width w/2.”A collection of split modes, where each split mode defines a single way to decompose a first block into multiple second blocks.
Single boundary of the picture
(Claim 1, Claim 4, Claim 7, Claim 10, Claim 13, Claim 15, Claim 17, Claim 19)		“A decoding method for decoding a current block, the current block including at least a portion of a picture, is disclosed that comprises: determining whether the current block extends beyond a boundary of the picture; determining for each split mode of a plurality of split modes whether the split mode is allowed for the current block by checking whether at least one of the split lines is co-located with one of the picture boundaries or whether a size of a block part, i.e. a remaining part resulting from the splitting, inside the picture along the picture boundary is a multiple of a minimum block size; decoding from a bitstream a current split mode of the current block responsive to the allowed split modes; and decoding the current block according to the current split mode.”One edge of the picture that the current block extends beyond.
Split direction is parallel
(Claim 1, Claim 4, Claim 7, Claim 10, Claim 13, Claim 15, Claim 17, Claim 19)		“According to a specific embodiment, the plurality of split modes comprises at least two of the following split modes, herein h is the height of the current block and w is the width of the current block: splitting the block horizontally into one upper sub-block of height h/4 and one bottom sub-block of height 3h/4; splitting the block horizontally into one bottom sub-block of height h/4 and one upper sub-block of height 3h/4; splitting the block vertically into one left sub-block of width w/4 and one right sub-block of width 3w/4; splitting the block vertically into one right sub-block of width w/4 and one left sub-block of width 3w/4; splitting the block horizontally into one sub-block of height h/4, one central sub-block of height h/2 and one sub-block of height h/4; splitting the block vertically into one sub-block of width w/4, one central sub-block of width w/2 and one sub-block of width w/4; splitting the block horizontally into two sub-blocks of height h/2; and splitting the block vertically into two sub-blocks of width w/2.”The direction of the split is oriented in the same direction as the single boundary of the picture.

Patent Family

Patent Family

File Wrapper

The dossier documents provide a comprehensive record of the patent's prosecution history - including filings, correspondence, and decisions made by patent offices - and are crucial for understanding the patent's legal journey and any challenges it may have faced during examination.

  • Date

    Description

  • Get instant alerts for new documents

US11917146

THOMSON LICENSING
Application Number
US16498392
Filing Date
Mar 13, 2018
Status
Granted
Expiry Date
Mar 13, 2038
External Links
Slate, USPTO, Google Patents