Two-Candidate Intra Prediction Mode Coding for Shorter Codewords
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing demand for high-resolution and high-quality video data leads to higher transmission and storage costs due to increased data volume, necessitating more efficient video compression techniques.
Innovation Solution
An intra-prediction mode decoding method using two candidate intra-prediction modes to enhance video coding efficiency by determining and decoding the intra-prediction mode of a current prediction unit based on fixed candidate modes, with additional bit information for mode identification and codeword mapping.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional intra-prediction mode encoding is used, then all possible modes can be represented, but the codeword length increases and coding efficiency decreases
Solution Approach 1:
The patent segments the intra-prediction mode encoding into two parts: first determining whether the mode is among the top 2 candidates using 1 bit, then only encoding the difference from the most likely candidate. This segmentation allows representing all modes while using shorter codewords on average, resolving the contradiction between coding efficiency and codeword length.
Solution Approach 2:
The patent performs preliminary action by pre-determining the top 2 candidate modes based on neighbor PU information before actual encoding. This preliminary selection allows the encoder to use a simplified encoding scheme that only needs to represent the difference from the most likely candidate, reducing the average codeword length while maintaining the ability to represent all possible modes.
2Measurement precision
If more candidate modes are used for intra-prediction, then prediction accuracy improves, but encoding complexity increases
Solution Approach 1:
The patent extracts only the most relevant information needed for encoding by focusing only on the top 2 candidate modes derived from neighbor PU data. Instead of encoding all possible modes or using complex multi-mode prediction, the invention extracts and uses only the essential candidate information, achieving good prediction accuracy with simple encoding.
Solution Approach 2:
The patent uses a simple and lightweight encoding approach that only needs to represent whether the mode is one of the top 2 candidates and which one it is. This 'cheap' encoding method using minimal bits achieves effective prediction without the complexity of encoding all possible modes or using complex prediction algorithms.
Data Source
AI summary
The method for decoding an intra-picture prediction mode includes the steps of: determining whether the intra-picture prediction mode of a current prediction unit is identical to a first intra-picture prediction mode candidate or a second intra-picture prediction mode candidate based on bit information; and when the intra-picture prediction mode of the current prediction unit is identical to the first intra-picture prediction mode candidate and/or to the second intra-picture prediction mode candidate, determining whether the first intra-picture prediction mode candidate or the second intra-picture prediction mode candidate is identical to the intra-picture prediction mode of the current prediction unit on the basis of additional bit information, and decoding the intra-picture prediction mode of the current prediction unit.