Nonrectangular Block Splitting for Higher-Efficiency Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in achieving high coding efficiency and image quality due to limitations in block splitting structures, particularly with the restriction that blocks split into binary or ternary trees cannot be further split into quadtrees, leading to inefficiencies in prediction and transform processes.
Innovation Solution
The introduction of a nonrectangular block splitting structure, which allows for additional splitting methods beyond traditional quadtree, binary tree, and ternary tree structures, enabling more effective prediction and transform processes by generating reference samples and applying nonrectangular splitting flags and indices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional quadtree, binary tree, and ternary tree splitting structures are used, then the block splitting process is simple and straightforward, but the coding efficiency and video quality are limited due to the restriction that blocks split into binary or ternary trees cannot be further split into quadtrees
Solution Approach 1:
The patent divides the block splitting process into multiple stages with different splitting modes (quadtree, binary tree, ternary tree, and nonrectangular splitting). Each stage can be independently selected and applied, allowing the system to segment the complex splitting decision into manageable parts while achieving higher coding efficiency through the nonrectangular splitting option.
Solution Approach 2:
The patent introduces dynamic flexibility by allowing the switching between different splitting structures based on the nonrectangular splitting flag. The block splitting structure is no longer fixed but can adaptively change between traditional rectangular splitting and nonrectangular splitting, enabling optimal performance for different block types and content characteristics.
2Manufacturing precision
If blocks are restricted to rectangular splitting only, then the prediction and transform processes are simpler, but the video quality and coding efficiency cannot be sufficiently improved
Solution Approach 1:
The patent introduces nonrectangular splitting structures that break the traditional symmetric rectangular block division. By allowing asymmetric and irregular block shapes in the splitting process, the system can better adapt to various video content characteristics, improving prediction accuracy and transform efficiency while enhancing overall video quality.
Solution Approach 2:
The patent changes the fundamental parameter of block shape from fixed rectangular to variable nonrectangular forms. This parameter change enables the block splitting structure to better match the actual content characteristics in different regions of the video, improving both prediction and transform processes through more appropriate block geometries.
3Adaptability or versatility
If multiple splitting structures (quadtree, binary tree, ternary tree) are supported, then the adaptability to different block types is improved, but the complexity of the splitting process increases due to the restriction on subsequent splitting operations
Solution Approach 1:
The patent creates a universal block splitting framework that can handle multiple splitting modes (quadtree, binary tree, ternary tree, and nonrectangular splitting) within a single unified process. The nonrectangular splitting flag serves as a universal control mechanism that can be applied at any stage of the splitting process, making the system multi-functional while managing complexity through standardized control logic.
Data Source
AI summary
A method and an apparatus are disclosed for video coding using a nonrectangular block splitting structure prediction and transform accordingly. The video coding method and the apparatus additionally apply a nonrectangular block splitting structure, in addition to an existing splitting structure, to a current block. The video coding method and the apparatus generate reference samples according to the applied splitting structure and effectively perform prediction and transform on split blocks.