Subpixel Interpolation Filter Coefficients for Video Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding technologies face challenges in achieving high compression ratios with minimal sacrifice in picture quality, particularly in streaming and storage due to limited bandwidth and memory resources.
Innovation Solution
The method involves using a subpixel interpolation filter with filter coefficients that satisfy a linear combination of a fractional sample position and a strength parameter, allowing for accurate intra- and inter-prediction processing by calculating coefficients on-the-fly without the need for look-up tables, thereby reducing memory demands and latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional interpolation filters with fixed coefficients are used, then implementation is simple, but prediction accuracy at subpixel positions deteriorates
Solution Approach 1:
The patent implements dynamic filter coefficients that adapt to subpixel positions. The filter transitions from static fixed coefficients to dynamic coefficients calculated based on fractional sample positions, allowing the filter to optimize prediction accuracy for each specific subpixel location while maintaining a manageable implementation through systematic calculation methods.
Solution Approach 2:
The patent changes the parameters of the interpolation filter by introducing filter strength parameters and phase parameters that vary with subpixel position. This allows the filter characteristics to be adjusted dynamically based on the fractional sample position, improving prediction accuracy without requiring completely different filter structures for each position.
2Measurement precision
If multiple interpolation filters are used for different subpixel positions, then prediction accuracy improves, but memory requirements and processing complexity increase
Solution Approach 1:
The patent creates a universal interpolation filter that can handle all subpixel positions through dynamic parameter adjustment rather than requiring separate dedicated filters for each position. The same filter structure serves multiple functions by adapting its coefficients based on the input subpixel position, reducing memory requirements while maintaining accuracy.
Solution Approach 2:
Instead of storing multiple static filter sets in memory, the patent implements a dynamic coefficient generation system that calculates appropriate filter coefficients on-demand based on the fractional sample position. This eliminates the need for large lookup tables or multiple stored filter configurations.
3Manufacturing precision
If complex interpolation algorithms are used to achieve high prediction accuracy, then picture quality improves, but processing time and computational load increase
Solution Approach 1:
The patent achieves high prediction accuracy through efficient parameter-based coefficient calculation rather than complex iterative algorithms. By expressing filter coefficients as functions of simple parameters like fractional sample positions and filter strength, the system maintains high picture quality while reducing computational complexity and processing time.
4Speed
If filter coefficients are pre-calculated and stored, then processing speed improves, but adaptability to different subpixel positions deteriorates
Solution Approach 1:
The patent implements a dynamic coefficient system where filter coefficients are generated on-demand based on the specific subpixel position requirements. This maintains adaptability to any subpixel position while achieving processing speed through efficient calculation formulas rather than slow lookup operations.
Solution Approach 2:
The filter system serves itself by automatically generating appropriate coefficients based on the input position parameters without requiring external lookup tables or pre-computed data. The system uses the fractional sample position information already available in the decoding process to self-generate the necessary filter coefficients.
Data Source
AI summary
The present disclosure relates to apparatuses and methods for intra- or inter-prediction for video encoding and decoding. In one embodiment, a method includes obtaining a reference sample and obtain a fractional sample position p; filtering the reference sample using a subpixel interpolation filter to obtain a predicted sample value. In the method, the filter coefficients of the subpixel interpolation filter satisfy a linear combination of p and a filter strength parameter S.