Image Coding Quantization Control for Rectangular Sub-Blocks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge in the VVC coding system is the inability to uniquely determine quantization parameters when rectangular sub-block splitting is performed, as the existing methods are based on square sub-block assumptions.
Innovation Solution
An image coding apparatus and decoding apparatus that split images into sub-blocks, determining quantization parameters based on the smaller size of the sub-block's horizontal or vertical dimensions relative to a quantization control size, allowing for appropriate encoding and decoding of quantization parameters for both square and rectangular sub-blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If rectangular sub-block splitting is performed in VVC, then coding efficiency is improved, but quantization parameter determination becomes ambiguous
Solution Approach 1:
The patent changes the parameter for determining quantization control from area-based (used in HEVC) to dimension-based (minimum of width or height). This parameter change allows rectangular sub-blocks to be handled consistently, as the minimum dimension provides a clear criterion for quantization parameter determination regardless of block shape, thereby resolving the ambiguity while maintaining coding efficiency improvements from rectangular splitting.
2Manufacturing precision
If quantization parameters are encoded for each sub-block, then image quality is improved, but data amount increases
Solution Approach 1:
The patent applies local quality by encoding quantization parameters selectively based on sub-block characteristics. Specifically, when the minimum dimension of a sub-block is greater than or equal to the quantization control size, a quantization parameter is encoded; otherwise, it is inferred from neighboring blocks. This selective approach ensures high image quality for larger sub-blocks while reducing data amount for smaller sub-blocks.
Solution Approach 2:
The patent implements partial action by not encoding quantization parameters for all sub-blocks uniformly. Instead, it encodes parameters only for sub-blocks meeting the size criterion (minimum dimension ≥ quantization control size), and infers parameters for smaller sub-blocks. This partial encoding approach optimizes the balance between image quality and data compression efficiency.
Data Source
AI summary
Control over encoding of a quantization parameter is appropriately enabled with not only square sub-blocks but also rectangular sub-blocks by using a quantization control size adaptively according to a shape of sub-blocks, with the result that coding efficiency is improved.