Adaptive Sub-block Scan Order for Video Coding Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding techniques face inefficiencies in scanning and coding transform coefficients, particularly in grouping non-zero coefficients together for effective entropy coding, which affects compression efficiency.
Innovation Solution
The proposed method involves dividing transform coefficients into sub-blocks and adapting the coefficient scan order and sub-block scan order to match each other, allowing for efficient scanning and coding by grouping non-zero coefficients together, and modifying the context support neighborhood to reduce dependencies and enhance parallelization during entropy coding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If transform coefficients are scanned using traditional methods, then the scanning process is simple, but non-zero coefficients are not effectively grouped together, reducing compression efficiency
Solution Approach 1:
The transform coefficient block is divided into multiple sub-blocks, and each sub-block is scanned separately using the same scan order. This segmentation allows non-zero coefficients to be grouped together more effectively within each sub-block, improving compression efficiency while maintaining a systematic and manageable scanning process.
Solution Approach 2:
The patent introduces adaptive scan order selection where the scan order (e.g., diagonal, vertical, horizontal) is dynamically chosen based on the characteristics of the transform coefficient block. This dynamic adaptation allows the scanning process to optimize coefficient grouping for different types of content, improving compression efficiency without requiring a completely complex fixed structure.
2Productivity
If the coefficient scan order and sub-block scan order are matched, then non-zero coefficients are grouped effectively, but the coding process becomes more complex
Solution Approach 1:
The patent applies different scan orders to different sub-blocks based on their local characteristics. Each sub-block can use the scan order that best groups its non-zero coefficients, allowing localized optimization of coding efficiency without requiring a completely complex global scanning strategy.
Solution Approach 2:
The scan order is determined and established before the actual coefficient scanning and coding process begins. This preliminary determination of scan orders for each sub-block simplifies the subsequent coding process by pre-organizing the scanning strategy, reducing the complexity during the actual coding phase.
3Speed
If dependencies in entropy coding are reduced, then parallel processing is enhanced, but the context model becomes more complex
Solution Approach 1:
The entropy coding process is segmented into independent parallel contexts, where each sub-block's coefficient coding can proceed independently with its own context model. This segmentation reduces dependencies between different parts of the coding process, enabling parallel processing while managing context model complexity through modular organization.
Solution Approach 2:
The patent introduces an intermediate scan order determination step that acts as a mediator between the transform coefficients and the entropy coding process. This intermediate step organizes the coefficients in a way that reduces dependencies for the subsequent entropy coding, facilitating parallel processing while keeping the context model manageable.
Data Source
AI summary
In an example, aspects of this disclosure relate to a method for decoding transform coefficients in a video decoding process that includes decoding a one-dimensional array of transform coefficients, and performing a scan on the one-dimensional array of transform coefficients according to a sub-block scan order and a coefficient scan order to produce a block of transform coefficients, where the sub-block scan order comprises an order in which each sub-block of a plurality of sub-blocks within the block is scanned, and where the coefficient scan order comprises an order in which the transform coefficients corresponding to each sub-block of the plurality of sub-blocks are scanned.