Adaptive Block Division Modes for Video Texture Fitting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies have limited block division methods, leading to inefficient fitting of image texture, which affects the overall efficiency of video coding due to the use of regular, straight-line boundaries in sub-block divisions.
Innovation Solution
A method and system for block division in video processing that includes multiple division modes such as equal, derivative, combination, and step division modes, allowing for flexible block division based on image texture, with the target division mode determined by comparing costs and suitability for each block, enabling adaptive sub-block creation and encoding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If regular straight-line block division is used, then the block division structure is simple, but the fitting performance to image texture deteriorates
Solution Approach 1:
The patent segments the block division process into multiple modes (equal division, derivative division, combination division, step division) with different boundary characteristics. This allows selection of appropriate segmentation patterns based on image texture requirements, resolving the contradiction between structural simplicity and fitting precision.
Solution Approach 2:
The patent introduces dynamic block division boundaries that can adapt to image content characteristics. By using step division modes and derivative division modes, the boundaries become flexible rather than fixed straight lines, enabling better adaptation to texture variations while maintaining manageable structural complexity.
2Manufacturing precision
If multiple complex division modes are introduced, then the fitting performance to image texture improves, but the device complexity increases
Solution Approach 1:
The patent implements a hierarchical division strategy where not all blocks require complex division modes. By applying sophisticated division methods (step division, combination division) only where needed based on image characteristics, the system achieves high fitting performance without uniformly increasing complexity across all blocks.
Solution Approach 2:
The patent creates a universal block division framework that can handle multiple image types and texture patterns through a set of standardized division modes. This multi-functional approach allows the same system structure to adapt to various scenarios, achieving high precision without proportionally increasing overall system complexity.
3Manufacturing precision
If the division depth is increased, then the fitting performance improves, but the encoding efficiency deteriorates
Solution Approach 1:
The patent applies different division depths and modes locally based on image texture characteristics. By using step division and derivative division modes that can achieve better fitting with fewer levels, the system reduces the need for deep recursive division, thereby improving encoding efficiency while maintaining fitting performance where it matters most.
Data Source
AI summary
The present disclosure relates to systems and methods for block division and encoding in video processing. A method may include obtaining a block to be predicted in an image frame. The method may also include determining, based on a plurality of division modes of the block, a target division mode of the block. The plurality of division modes may include an equal division mode that is configured to divide the block into multiple sub-blocks with a same size. The method may further include dividing, based on the target division mode, the block into a plurality of sub-blocks, and encoding the block based on the plurality of sub-blocks.