Intra Prediction Using Extended Boundary Pixels for SIMD Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding techniques face inefficiencies in intra-frame prediction due to complex decision processes and computational bottlenecks, particularly in selecting reference pixels and calculating intersection positions, which hinder implementation on Single-Instruction Multiple Data (SIMD) architectures.
Innovation Solution
The proposed solution simplifies the intra prediction process by eliminating the need to select between horizontal and vertical boundary pixels and reduces recurring calculations, allowing for the extension of pixel arrays using retrieved vertical and horizontal boundary pixels, thereby enabling efficient implementation on SIMD architectures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If the traditional intra prediction process is used with decision processes for selecting boundary pixels and calculating intersection positions, then prediction accuracy is maintained, but computational complexity increases and implementation on SIMD architectures becomes difficult
Solution Approach 1:
The patent extracts and eliminates the complex decision-making processes and intersection position calculations from the intra prediction algorithm. By removing these computationally intensive components and replacing them with a simplified approach that uses only horizontal boundary pixels, the patent reduces computational complexity while preserving essential prediction functionality.
Solution Approach 2:
The patent segments the prediction process into simpler, independent operations that can be executed in parallel. By dividing the target block into multiple prediction blocks and applying simplified prediction modes independently to each segment, the patent enables SIMD architecture implementation while maintaining overall prediction accuracy.
2Productivity
If divisional operations and branching processes are used in intra prediction, then accurate prediction directions are achieved, but processing speed decreases and parallel processing becomes inefficient
Solution Approach 1:
The patent replaces the mechanical calculation processes (divisional operations and branching logic) with a simpler data access pattern. Instead of calculating intersection positions through division and making branching decisions, the patent directly accesses horizontal boundary pixels using a unified formula, eliminating computational overhead and enabling efficient parallel processing on SIMD architectures.
3Measurement precision
If multiple prediction modes with different boundary selections are implemented, then prediction quality is improved, but device complexity and operational difficulty increase
Solution Approach 1:
The patent creates a universal prediction approach that works for all prediction directions using the same horizontal boundary pixel access method. This unified formulation eliminates the need for separate handling of different boundary selection cases, making the implementation easier while maintaining prediction quality through the consistent application of the simplified model across all modes.
Data Source
Figure 1
Figure 2~3
Figure 4
AI summary
The present invention provides a unique intra prediction process which improves the efficiency of video coding. H.264/AVC uses reference pixels in a horizontal boundary located immediately above a target block to be predicted and reference pixels in a vertical boundary located immediately left of the target block. In the present invention, at least some of one of an array of horizontal boundary pixels and an array of vertical boundary pixels are retrieved. Then, the retrieved pixels are added to the other boundary pixels to extend the array thereof. Intra prediction is performed, based solely on the extended array of boundarypixels.