Adaptive Loop Filter Virtual Boundary Processing for 360 Video
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In video coding systems, particularly in the emerging Versatile Video Coding (VVC) standard, applying in-loop filters across discontinuous edges in 360° videos results in poor visual quality and decreased coding efficiency due to the processing of uncorrelated pixels across these edges.
Innovation Solution
The proposed solution involves treating discontinuous edges as virtual boundaries in the Adaptive Loop Filter (ALF) processing, allowing in-loop filtering to be applied while avoiding the use of uncorrelated samples. This is achieved through virtual boundary processing, where unavailable samples are padded, and by prioritizing horizontal virtual boundaries over vertical ones in certain cases.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If in-loop filters are applied across discontinuous edges in 360° videos, then filtering coverage is improved, but visual quality deteriorates due to processing of uncorrelated pixels
Solution Approach 1:
The patent segments the picture into multiple sub-pictures and introduces virtual boundaries to divide the filtering process. By treating discontinuous edges as virtual boundaries, the filter processing is segmented into regions that respect the discontinuities, preventing uncorrelated pixels from being processed together while maintaining adequate filtering coverage within each segment.
Solution Approach 2:
The patent applies different filtering strategies to different regions of the picture. At virtual boundaries corresponding to discontinuous edges, the filtering is modified to use only correlated samples from the same sub-picture, while in regular regions standard filtering applies. This local differentiation maintains visual quality at boundaries while preserving filtering benefits in continuous regions.
2Manufacturing precision
If virtual boundary processing is applied to avoid uncorrelated samples, then visual quality is improved, but device complexity increases due to additional boundary detection and sample padding operations
Solution Approach 1:
The patent performs preliminary detection of virtual boundaries and pre-pads unavailable samples before the actual filtering process. By preparing the boundary conditions and padding samples in advance, the main filtering operation can proceed without complex runtime checks, reducing processing complexity while maintaining visual quality improvements.
3Productivity
If horizontal virtual boundaries are prioritized over vertical ones, then processing efficiency is improved, but filtering accuracy may deteriorate in cases where vertical boundaries should have higher priority
Solution Approach 1:
The patent introduces a priority parameter system for different boundary orientations. Instead of fixed horizontal priority, the system dynamically determines boundary priority based on the specific context and discontinuity characteristics. This allows the filtering accuracy to be maintained by adapting to the actual picture content while still providing efficient processing through defined priority rules.
Data Source
AI summary
Methods and apparatus for video coding using Adaptive Loop Filter (ALF) processing are disclosed. According to one method, the ALF virtual boundary processing is always applied to the bottom CTU row regardless whether the bottom CTU row is for a picture or a sub-picture. In another method, the ALF padding process is unified for different boundary types belonging to the target boundary belongs to a boundary-type group comprising two or more of slice boundary, tile boundary, VR360 face boundary, and sub-picture boundary. In yet another method, the ALF VB processing has a fixed priority for a corner region that both horizontal and vertical virtual boundaries can be applied.