Prediction Block Boundary Smoothing Using Dominant Direction Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding standards like VVC and ECM lack effective methods for smoothing the boundaries of prediction blocks, particularly in intra and inter coding modes, leading to suboptimal prediction block boundaries.
Innovation Solution
A method is introduced to smooth prediction block boundaries by analyzing a reference area to determine one or more dominant directions, adjusting pixel values along these directions using statistical or searching methods, and applying a smoothing process to align boundary pixels with neighboring reconstructed blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If no boundary smoothing is applied to prediction blocks, then the coding process remains simple and fast, but the video quality and compression efficiency deteriorate due to misaligned block boundaries
Solution Approach 1:
The patent applies boundary smoothing as a preliminary action before final prediction block reconstruction. By pre-smoothing the boundaries of neighboring blocks and the current prediction block using gradient analysis and pixel value adjustments, the method ensures better boundary alignment is established before decoding, thereby improving video quality without adding complex post-processing steps
Solution Approach 2:
The patent applies local quality by performing boundary smoothing selectively on boundary pixels rather than the entire prediction block. The smoothing operation is localized to pixels adjacent to block boundaries, using gradient calculations and weighted adjustments only where boundary misalignment occurs, thus improving precision without proportionally increasing overall complexity
2Reliability
If directional boundary smoothing is applied, then video quality and compression efficiency improve, but computational complexity and processing time increase
Solution Approach 1:
The patent applies partial action by performing boundary smoothing on only the necessary boundary pixels rather than the entire prediction block. The gradient calculation and pixel adjustment operations are limited to pixels at block boundaries that require smoothing, reducing the overall computational burden while maintaining prediction accuracy for the most critical regions
Solution Approach 2:
The patent segments the boundary smoothing process into distinct directional components (horizontal, vertical, and diagonal gradients). By analyzing and smoothing boundaries in separate directional passes, the method improves prediction accuracy for each orientation while organizing computations in a modular fashion that can be efficiently implemented and optimized
Data Source
AI summary
An apparatus configured to: analyze a reference area of a prediction block to determine at least one dominant direction of the reference area of the prediction block; and adjust a pixel value of a boundary pixel of the prediction block along the at least one dominant direction of the reference area of the prediction block.