Adaptive In-Loop Filtering for Lower-Complexity Video Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video compression techniques face inefficiencies in making optimal in-loop filtering decisions, particularly with the sequential application of SAO and ALF filters, leading to wasted bits and increased complexity without achieving optimal quality improvements.
Innovation Solution
A single pass encoding solution that adaptively disables SAO filtering based on frame type and quantization parameters, and makes ALF decisions after SAO, optimizing in-loop filtering to reduce bitrate and complexity while maintaining quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If SAO and ALF filters are applied sequentially with full RDO decisions, then filtering quality is improved, but bitrate and processing complexity increase
Solution Approach 1:
The patent applies partial action by making ALF RDO decisions only for certain blocks (e.g., luma blocks) while skipping SAO RDO decisions for other blocks (e.g., chroma blocks), rather than applying full RDO decisions to all blocks. This selective approach reduces processing complexity and bitrate while maintaining filtering quality for the most important components.
Solution Approach 2:
The patent applies local quality by making different filtering decisions for different block types and locations. Specifically, it makes ALF RDO decisions for luma blocks but skips SAO RDO decisions for chroma blocks, recognizing that luma blocks contribute more to perceived video quality. This localized optimization reduces overall complexity while preserving quality where it matters most.
2Manufacturing precision
If SAO RDO decisions are made for all blocks, then filtering optimization is improved, but bitrate increases due to unnecessary SAO usage
Solution Approach 1:
The patent applies partial action by making ALF RDO decisions only for certain blocks (e.g，luma blocks) while skipping SAO RDO decisions for other blocks (e.g., chroma blocks), rather than applying full RDO decisions to all blocks. This selective approach reduces processing complexity and bitrate while maintaining filtering quality for the most important components.
Solution Approach 2:
The patent extracts and removes unnecessary SAO RDO decisions for chroma blocks, recognizing that these decisions would have minimal impact on overall video quality while consuming valuable bitrate. By taking out these unnecessary decisions, the patent reduces bitrate without significantly compromising filtering optimization.
3Manufacturing precision
If adaptive in-loop filtering is applied to all frames, then quality improvement is maximized, but processing time increases
Solution Approach 1:
The patent applies partial action by making ALF RDO decisions only for certain blocks (e.g., luma blocks) while skipping SAO RDO decisions for other blocks (e.g., chroma blocks), rather than applying full RDO decisions to all blocks. This selective approach reduces processing complexity and bitrate while maintaining filtering quality for the most important components.
Solution Approach 2:
The patent segments the filtering process into different parts, making ALF decisions for luma blocks but skipping SAO decisions for chroma blocks. This segmentation allows the system to focus computational resources on the most time-critical and quality-impactful operations, reducing overall processing time while maintaining acceptable quality improvement.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
In some compression techniques, in-loop filters can be included to effectively remove coding artifacts and improve the objective quality measurement at the same time. To avoid increasing complexity in the encoder, a single pass encoding solution can be implemented to efficiently and effectively make reasonably optimal in-loop filtering decisions. The solution can improve the in-loop filtering bit usage (e.g., reduce bitrate) and reduce the complexity in the encoder at the same time.