Scaling List Adaptation for HEVC Quantization Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In HEVC, the uniform scaling list applied during quantization leads to reduced coding efficiency due to the use of scaling lists similar to those for small-size TUs on large-size TUs, which has a minimal impact on subjective image quality but can decrease coding efficiency.
Innovation Solution
An image processing apparatus that sets a scaling list with values appropriate to the current block size, allowing larger values for higher frequency components in larger blocks and smaller values for lower frequency components in larger blocks, and includes a flatness detection section to set larger blocks in flat portions, ensuring optimal coding efficiency while maintaining subjective image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a uniform scaling list is applied to all TU sizes in HEVC quantization, then the implementation is simple and consistent, but coding efficiency is reduced due to inappropriate scaling for large-size TUs
Solution Approach 1:
The patent applies different scaling list values to different frequency components based on TU size. For large-size TUs (16×16, 32×32), larger scaling values are applied to high-frequency components while smaller values are applied to low-frequency components, rather than using a uniform scaling list for all frequencies and all TU sizes.
Solution Approach 2:
The patent changes the scaling list parameters based on the TU size. Different scaling list configurations are selected according to the specific TU size (4×4, 8×8, 16×16, 32×32), allowing the quantization process to adapt to the characteristics of each block size and improve coding efficiency.
2Productivity
If larger block sizes are used in flat portions to improve coding efficiency, then the compression ratio increases, but the scaling list must be appropriately adjusted to maintain image quality
Solution Approach 1:
The patent detects flat portions in the image and applies appropriate scaling list values specifically to large-size TUs in these regions. The scaling list is configured with larger values for high-frequency components in flat portions, which allows aggressive compression while maintaining perceived image quality since human vision is less sensitive to high-frequency details in uniform regions.
Data Source
AI summary
The present disclosure relates to an image processing apparatus and method that can keep reduction in coding efficiency to a minimum. A scaling list is set that has values appropriate to a current block size for the current block to be quantized in an image to be coded, quantization is performed using the set scaling list, and coded data is generated by coding an acquired quantization coefficient. The present disclosure is applicable, for example, to an image processing apparatus, an image coding apparatus, an image decoding apparatus, and so on.