3-Tap Directional Reference Filtering for Intra Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image compression techniques face challenges in enhancing prediction efficiency and reducing artifacts in intra prediction processes, particularly in high-resolution and high-quality images, without increasing complexity.
Innovation Solution
An intra prediction method that filters neighboring reference samples based on the prediction mode and block size, using a 3-tap filter with coefficients [1 2 1] to generate predicted samples, and adaptively determines whether filtering is necessary.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If filtering is applied to all neighboring reference samples in intra prediction, then prediction accuracy improves, but computational complexity increases
Solution Approach 1:
The patent applies filtering selectively to specific neighboring reference samples based on the intra prediction mode. For directional prediction modes, only reference samples in the prediction direction and its immediate neighbors are filtered, while other reference samples remain unfiltered. This localized filtering approach maintains prediction accuracy for critical samples while reducing overall computational complexity.
Solution Approach 2:
The patent implements partial filtering by applying the filtering operation to only a subset of neighboring reference samples rather than all samples. Specifically, it filters the reference sample in the prediction direction and a limited number of adjacent samples, which is sufficient to improve prediction accuracy without the excessive computational cost of filtering all available reference samples.
2Manufacturing precision
If high-resolution images are processed with existing compression techniques, then image quality is maintained, but information transfer cost and storage cost increase
Solution Approach 1:
The patent changes the parameter of reference sample processing by introducing selective filtering based on prediction mode. This modification improves the accuracy of intra prediction, which in turn enhances compression efficiency. The improved compression efficiency reduces the number of bits required to represent high-resolution images, thereby lowering both transfer cost and storage cost while maintaining image quality.
3Manufacturing precision
If filtering is applied to reduce artifacts in reconstructed images, then image quality improves, but processing time increases
Solution Approach 1:
The patent applies filtering locally to only those reference samples that most contribute to prediction accuracy, rather than uniformly filtering all samples. This selective approach reduces the total number of filtering operations required, thereby reducing processing time while still effectively reducing artifacts in the reconstructed image through targeted filtering of critical samples.
Data Source
AI summary
An intra prediction method and a device using the intra prediction method are provided. The intra prediction method includes the steps of: deriving a current prediction mode as a prediction mode of a current block; constructing neighboring samples of the current block with available reference samples; filtering the available reference samples; and generating predicted samples of the current block on the basis of the filtered available reference samples,. The filtering step includes performing the filtering using the available reference sample located in the prediction direction of the current prediction mode and a predetermined number of available reference samples neighboring to the prediction direction of the current prediction mode.