Implicit TU Boundary Deblocking for VVC Blocking Artifacts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The current draft of the H.266 (VVC) specification does not account for blocking artifacts that can occur at implicit transform unit (TU) boundaries, which are not explicitly addressed in the deblocking process.
Innovation Solution
A method is introduced to apply deblocking filters on implicit vertical and horizontal TU boundaries by splitting coding units (CUs) into smaller blocks without explicit syntax elements, using a maximum size threshold to determine implicit splits, and applying deblocking filters to the boundaries between these blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If deblocking filter is applied only on explicit TU boundaries as per current VVC specification, then device complexity is reduced, but blocking artifacts appear at implicit TU boundaries degrading video quality
Solution Approach 1:
The patent segments the TU boundary treatment into two categories: explicit boundaries (currently handled) and implicit boundaries (newly addressed). By dividing the boundary types and applying appropriate deblocking strategies to each, the patent resolves the contradiction by extending deblocking to implicit boundaries without requiring complete redesign of the deblocking filter architecture.
Solution Approach 2:
The patent applies local quality by treating implicit and explicit boundaries differently. For implicit boundaries, it uses gradient calculations and selective filtering based on local characteristics (intra/inter prediction modes, transform coefficients), while maintaining the existing deblocking approach for explicit boundaries. This localized adaptation reduces overall complexity while eliminating artifacts at critical implicit boundaries.
2Object-affected harmful factors
If deblocking filter is applied on all TU boundaries including implicit ones, then video quality is improved, but processing time and computational load increase
Solution Approach 1:
The patent implements partial action by applying deblocking filtering selectively rather than uniformly to all implicit boundaries. It uses gradient calculations and boundary strength parameters to determine which implicit boundaries require filtering, applying the filter only where necessary to reduce artifacts. This approach improves quality at affected boundaries while avoiding unnecessary processing at boundaries where artifacts are not present.
Solution Approach 2:
The patent changes parameters by introducing boundary strength (bs) values and gradient thresholds that control the intensity and application of deblocking filters. By adjusting these parameters based on local image characteristics (prediction modes, transform coefficients, gradient magnitudes), the patent optimizes the balance between artifact reduction and processing efficiency, applying stronger filtering only where needed.
3Ease of operation
If implicit TU boundaries are not deblocked, then encoding/decoding simplicity is maintained, but discontinuities appear at block boundaries reducing subjective quality
Solution Approach 1:
The patent applies preliminary action by calculating gradients and determining boundary strength parameters before applying the deblocking filter at implicit boundaries. This pre-processing step identifies which boundaries require filtering and prepares the necessary parameters, allowing the actual filtering operation to proceed efficiently without complex decision-making during the filtering stage itself.
Solution Approach 2:
The patent implements self-service by using the existing deblocking filter infrastructure and parameter structures already present in the VVC specification. The implicit boundary deblocking reuses existing components (gradient calculations, boundary strength parameters, filter kernels) rather than introducing entirely new mechanisms, allowing the system to handle implicit boundaries using its own existing tools with minimal additional complexity.
Data Source
AI summary
In one aspect there is disclosed a method of applying deblocking on implicit vertical TU boundaries when the CU width is larger than the maximum TU width and applying deblocking on implicit horizontal TU boundaries when the CU height is larger than the maximum TU height. Some exemplary embodiments include HEVC deblocking and deblocking using longer filters.