Polynomial Prediction for Smooth Video Blocks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding standards struggle with efficiently encoding smooth surfaces, leading to unnecessary overhead and reduced quality, especially in high-definition and high dynamic range (HDR) video sequences.
Innovation Solution
The method involves determining parameters for a polynomial model to predict smooth surface blocks, adjusting the quantization parameter (QP) based on the smoothness of blocks, and using low-frequency base functions for accurate representation and prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If standard video coding methods are used for smooth surfaces, then encoding is simple, but quality deteriorates and bitrate overhead increases
Solution Approach 1:
The patent applies different encoding strategies to different regions of the image based on their characteristics. Smooth regions are identified using gradient analysis and processed with specialized polynomial-based prediction methods, while non-smooth regions use standard coding techniques. This local differentiation improves overall quality without unnecessarily increasing bitrate for all regions.
Solution Approach 2:
The patent changes the encoding parameters dynamically based on region smoothness. For smooth regions, it uses polynomial order parameters (e.g., first-order, second-order polynomials) and gradient-based prediction parameters instead of standard transform coding parameters. This parameter adaptation allows efficient representation of smooth surfaces with reduced bitrate overhead.
2Manufacturing precision
If quantization parameter is reduced for smooth blocks, then quality improves, but bitrate overhead increases
Solution Approach 1:
The patent applies gradient-based smoothness detection to identify regions where quality improvement is most beneficial. By calculating local gradients and comparing them to thresholds, the system determines which blocks are truly smooth and would benefit from reduced QP values, avoiding unnecessary bitrate increases in regions that don't require high quality.
Solution Approach 2:
Instead of reducing QP uniformly across all blocks, the patent applies partial action by selectively reducing QP only for blocks that pass the smoothness test. This prevents excessive bitrate consumption while still achieving quality improvements in the most critical smooth regions like skies and walls.
3Measurement precision
If polynomial model parameters are used for prediction, then smooth surface accuracy improves, but computational complexity increases
Solution Approach 1:
The patent segments the image processing task into distinct stages: gradient calculation, smoothness classification, polynomial parameter estimation, and prediction. Each stage uses optimized algorithms appropriate to its function, reducing overall computational complexity while maintaining prediction accuracy for smooth surfaces.
Solution Approach 2:
The patent applies polynomial models partially - only to blocks identified as smooth through gradient analysis. For non-smooth blocks, standard prediction methods are used. This partial application significantly reduces computational complexity while maintaining high prediction accuracy where it matters most.
Data Source
AI summary
A method for making a smooth prediction block of samples in a picture in an image or video encoder or decoder is provided. The method includes determining parameters r of a polynomial model by r=(BTB)−1*(BT*x), where B is a matrix of bases and x is a source of samples in vector form. The method includes predicting a block x′ based on the parameters r and the matrix of bases B.