In-Loop Filtering Across Virtual Boundaries in Image Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing demand for high-resolution, high-quality images and videos, such as 4K or 8K, leads to higher data transmission and storage costs, and existing compression technologies struggle to efficiently compress and transmit immersive media like VR and AR content effectively, particularly in signaling in-loop filtering-related information.
Innovation Solution
The implementation of a method and apparatus for efficient image/video coding that includes deblocking, sample adaptive offset (SAO), and adaptive loop filtering (ALF) based on virtual boundaries, with a sequence parameter set (SPS) flag enabling in-loop filtering across these boundaries, improving coding efficiency and visual quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-resolution, high-quality image/video compression is applied, then transmission and storage costs increase, but image quality is improved
Solution Approach 1:
The patent changes the parameter of filtering boundary definition by introducing virtual boundaries that differ from traditional block boundaries. This allows filtering to be applied more selectively and efficiently across picture partitions, improving compression efficiency while maintaining image quality through optimized in-loop filtering operations.
2Measurement precision
If in-loop filtering is applied across virtual boundaries, then visual quality is improved, but hardware resources are consumed
Solution Approach 1:
The patent introduces dynamic control of in-loop filtering through flags and parameters that allow the filtering operation to be adaptively applied or disabled across virtual boundaries. This dynamic approach enables the system to optimize hardware resource usage by applying filtering only when and where it provides visual quality benefits, rather than uniformly across all boundaries.
3Productivity
If filtering information is signaled efficiently, then coding efficiency is improved, but device complexity increases
Solution Approach 1:
The patent segments the filtering control information into multiple independent flags and parameters (e.g., sps_virtual_boundaries_enabled_flag, pic_virtual_boundaries_enabled_flag, rbp_virtual_boundaries_enabled_flag) that can be selectively applied to different picture regions and boundary types. This segmentation allows for efficient signaling by enabling fine-grained control without requiring complex data structures, as each flag independently controls filtering behavior for specific virtual boundary scenarios.
Data Source
AI summary
A decoding method performed by a decoding apparatus, according to embodiments of the present document, may comprise the steps of: acquiring, via a bitstream, image information including residual information; generating reconstructed samples of a current picture on the basis of the residual information; and generating modified reconstructed samples by performing an in-loop filtering procedure on the reconstructed samples of the current picture, wherein whether or not the in-loop filtering procedure is performed across virtual boundaries is determined. The virtual boundaries can bring, to the decoding apparatus, advantages from a hardware perspective.