Bitstream Decoder Combining Multiple Intra-Prediction Modes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image and video codecs using single intra-prediction modes may not provide optimal prediction of image block content, leading to suboptimal compression efficiency.
Innovation Solution
A method of decoding a bitstream that combines multiple intra-prediction hypotheses by determining and applying multiple intra-prediction modes to reference samples, forming a combined hypothesis block of predictions, which can include weighted averages or additions of prediction samples, to improve prediction accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single intra-prediction mode is used, then the decoding process is simple and fast, but the prediction accuracy and compression efficiency are suboptimal
Solution Approach 1:
The prediction process is segmented into multiple independent hypothesis generations (first hypothesis using first intra-prediction mode, second hypothesis using second intra-prediction mode), allowing parallel processing of different prediction modes before combining results
Solution Approach 2:
Multiple intra-prediction hypotheses are merged by computing a weighted average of prediction samples from different modes, combining the strengths of each mode to achieve superior prediction accuracy while maintaining a unified decoding process
2Measurement precision
If multiple intra-prediction modes are applied, then the prediction accuracy improves, but the computational complexity increases
Solution Approach 1:
Multiple intra-prediction hypotheses are generated in advance before the final prediction is needed, allowing the decoder to prepare multiple candidate predictions that can be quickly combined and compared against residual data
Solution Approach 2:
The system dynamically adjusts prediction parameters by selecting different intra-prediction modes and their corresponding weights based on the characteristics of each block, optimizing the balance between computational effort and prediction accuracy for different regions
Data Source
AI summary
The present disclosure relates to: a method of decoding, in a decoder, a bitstream representing an image, wherein the bitstream comprises a plurality of blocks of residual samples, and one or more reconstructed image frames of the image are obtained by adding a respective block of residual samples to a combined hypotheses block of intra-predictions, the combined hypotheses block of intra-predictions being formed by applying a method comprising: receiving the bitstream; for each block of residual samples: determining if a combined intra-prediction hypothesis is indicated and/or inferred from the bitstream and, if so: for each block of residual samples of a block for which a combined intra-prediction hypothesis is indicated and/or inferred: determining a first intra-prediction mode from information extracted from the bitstream; determining or inferring at least a second, different, intra-prediction mode; applying the first intra-prediction mode to at least one reference sample to form a first hypothesis prediction sample; applying the at least one second intra-prediction mode to said at least one reference sample to form a second hypothesis prediction sample; and forming a said combined hypotheses block of predictions by combining the first and second hypothesis prediction samples formed for the block.