Fractional Boundary Video Block Partitioning

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current video coding techniques face inefficiencies in partitioning fractional boundary video blocks, leading to a large number of small video blocks near picture boundaries, which adversely impacts coding efficiency due to increased syntax element transmission and parsing requirements.

Innovation Solution

The method involves partitioning fractional boundary video blocks using a subset of available partition modes, such as symmetric and asymmetric binary tree split modes, to minimize the impact on coding efficiency by applying predefined partitioning that aligns blocks with picture boundaries, thereby reducing the number of small blocks and associated syntax elements.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If conventional video coding techniques are used to partition video blocks, then complete coverage of picture boundaries is achieved, but the number of small video blocks increases significantly

Engineering Contradiction:

Improvepartitioning precisionVSAvoidnumber of video blocks

Core Design Contradiction:

Manufacturing precisionVSDevice complexity

Solution Approach 1:

The video block partitioning is segmented into two distinct types: conventional partitioning for blocks not at picture boundaries, and inferred partitioning for fractional boundary blocks. This segmentation allows each type to be handled with the most appropriate method, reducing the overall number of blocks while maintaining precision where needed.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Instead of applying conventional partitioning methods to all video blocks including fractional boundary blocks, the patent inverts the approach by applying inferred partitioning specifically to fractional boundary blocks. This inversion resolves the contradiction by using a specialized method for the problematic case rather than forcing a general method to fit all cases.

Inventive Principle:

Principle #13The other way round (Inversion)

2Adaptability or versatility

If conventional partitioning methods are applied to fractional boundary video blocks, then all blocks are treated uniformly, but syntax element transmission and parsing requirements increase

Engineering Contradiction:

Improvepartitioning flexibilityVSAvoidsyntax element data

Core Design Contradiction:

Adaptability or versatilityVSLoss of information

Solution Approach 1:

The patent applies local quality by treating fractional boundary blocks differently from conventional blocks. Instead of uniform treatment, the inferred partitioning method is applied locally to fractional boundary blocks based on their specific characteristics, reducing syntax element requirements for these particular blocks while maintaining appropriate flexibility where needed.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The patent changes the partitioning parameter for fractional boundary blocks from conventional methods to inferred partitioning. This parameter change reduces the syntax element transmission and parsing requirements specifically for fractional boundary blocks, addressing the information loss issue without sacrificing overall partitioning flexibility.

Inventive Principle:

Principle #35Parameter changes

3Adaptability or versatility

If the number of partition modes is increased to handle fractional boundary blocks, then partitioning options are expanded, but coding efficiency decreases

Engineering Contradiction:

Improvepartition mode optionsVSAvoidcoding efficiency

Core Design Contradiction:

Adaptability or versatilityVSProductivity

Solution Approach 1:

The patent applies partial action by using a subset of partitioning modes specifically for inferred partitioning of fractional boundary blocks, rather than expanding all available partition modes. This selective approach maintains coding efficiency by limiting the complexity increase to only what is necessary for handling fractional boundary cases.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS11259021B2Systems and methods for partitioning video blocks at a boundary of a picture for video coding

Publication Date: 2022.02.22 SHARP KK

AI summary

A method for determining a partitioning of video data for video coding is provided. A coding unit is determined from a picture of the video data. The coding unit is determined as a fractional boundary coding unit when the coding unit has only a portion within a boundary of the picture. A subset of partitioning modes is available for partitioning the fractional boundary coding unit and includes quadtree splitting and symmetric binary tree splitting partitioning modes. When the coding unit is the fractional boundary coding unit, the method determines whether one of a block height and a block width of the coding unit is greater than a corresponding one of threshold values. A determination of the partitioning includes determining whether one of symmetric binary tree splittings is allowed for the coding unit based on the one of the block height and the block width and the corresponding threshold value.