Intra Prediction Reference Sample Selection for Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The computational complexity of intra prediction in video coding and decoding increases with larger processing block sizes due to the need to average a larger number of reference samples for the DC coefficient.
Innovation Solution
A method and apparatus for intra prediction that select a second reference sample set from a first reference sample set based on a preset selection mode, such as a preset position, sampling rate, or statistical characteristics, to reduce the number of samples used for prediction, thereby reducing computational complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If all reference samples are used to calculate the DC coefficient through averaging, then the prediction accuracy is improved, but the computational complexity increases with larger processing block sizes
Solution Approach 1:
The patent extracts only the necessary reference samples (first and second reference samples from opposite sides) instead of using all reference samples. This selective extraction reduces the number of samples involved in the averaging calculation while still providing accurate prediction, thereby reducing computational complexity without significantly compromising prediction accuracy.
Solution Approach 2:
The patent segments the reference sample set into different groups (first reference samples from one side, second reference samples from the opposite side). By dividing the reference samples into segments and selectively using only certain segments for DC coefficient calculation, the computational burden is reduced while maintaining the essential prediction capability.
2Manufacturing precision
If a larger number of reference samples are used for DC coefficient calculation, then the prediction quality is improved, but the processing time increases
Solution Approach 1:
The patent extracts only the essential reference samples needed for accurate prediction (first and second reference samples from opposite sides of the current block). By extracting only these critical samples and excluding redundant ones, the processing time for calculating the DC coefficient is reduced while maintaining prediction quality.
3Measurement precision
If all neighboring reference samples are averaged to construct the DC coefficient, then the prediction accuracy is improved, but the number of calculations increases with block size
Solution Approach 1:
The patent extracts only the necessary reference samples (first reference sample from one side and second reference sample from the opposite side) instead of averaging all neighboring reference samples. This extraction approach reduces the number of calculations required while maintaining prediction accuracy, thereby improving coding and decoding efficiency.
Solution Approach 2:
The patent applies partial action by using only a subset of reference samples (specifically, samples from opposite sides) rather than all available reference samples. This partial usage is sufficient to achieve accurate prediction while significantly reducing the computational workload and improving processing efficiency.
Data Source
AI summary
Provided are an intra prediction method and apparatus, and a computer-readable storage medium. The method includes: acquiring a first reference sample set corresponding to the current processing block, wherein the first reference sample set includes at least one of a reference row adjacent to the current processing block and a reference column adjacent to the current processing block; selecting, based on a pre-set selection method, a second reference sample set from the first reference sample set, wherein the second reference sample set includes at least one reference sample in the reference row adjacent to the current processing block or the reference column adjacent to the current processing block; obtaining, based on the second reference sample set, a prediction value of the current processing block; and performing, based on the prediction value, intra prediction on the current processing block.