Intra Prediction MPM Lists Using Neighboring Angular Modes for Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in efficiently generating Most Probable Mode (MPM) lists for intra prediction mode coding, which heavily influence the efficiency of intra prediction processing, particularly in terms of processor load and compression ratio.
Innovation Solution
A method for constructing a Most Probable Mode (MPM) list by determining intra prediction modes based on the angular modes of neighboring blocks, ensuring commonly used modes are included, allowing for efficient coding by referencing indices rather than actual modes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If a conventional MPM list generation method is used, then the coding process is simpler, but the coding efficiency is lower and more bits are required
Solution Approach 1:
The patent applies preliminary action by pre-defining specific MPM list configurations based on the angular modes of neighboring blocks before coding the current block. The encoder determines the angular modes of left and above neighboring blocks, selects the appropriate pre-defined MPM list pattern, and populates the MPM list accordingly. This pre-planned approach ensures that commonly used intra prediction modes are included in the MPM list, reducing the bits required for coding without requiring complex real-time analysis during the coding process.
2Productivity
If the MPM list includes all possible intra prediction modes, then coding efficiency improves, but the list size and processing complexity increase
Solution Approach 1:
The patent applies local quality by creating different MPM list configurations tailored to specific local conditions - the angular modes of neighboring blocks. Instead of using a single universal MPM list, the encoder adapts the MPM list content based on the local characteristics of neighboring blocks. The method defines different patterns for different scenarios (e.g., when left block is angular mode LA and above block is angular mode AA, use pattern 1; when left block is Planar/DC and above block is angular mode, use pattern 2), ensuring that the MPM list size remains manageable while including the most relevant modes for each local context.
3Device complexity
If processor load is reduced by using simpler MPM list methods, then device complexity decreases, but compression ratio deteriorates
Solution Approach 1:
The patent applies parameter changes by modifying the MPM list generation parameters based on the angular modes of neighboring blocks. The method changes the composition and ordering of modes in the MPM list according to the specific configuration of neighboring blocks, rather than using a fixed, complex generation algorithm. This allows the system to maintain low processor load by using simple rule-based selection while improving compression ratio through adaptive optimization of the MPM list content for each coding scenario.
Data Source
Figure 1A
Figure 1B
Figure 2
AI summary
It is provided a method of prediction coding of a current block implemented by a decoding device or an encoding device, comprising: obtaining an intra prediction mode of a left neighbor block of the current block; obtaining an intra prediction mode of an above neighbor block of the current block; constructing a Most Probable Mode, MPM, list of intra prediction modes for the current block, the MPM list comprising at least 5 entries of intra prediction modes as follows: {ang, 2 + ( ( ang + 61) % 64 ), 2 + ((ang - 1 ) % 64 ), 2 + ( ( ang + 60 ) % 64 ), 2 + ((ang) % 64 )} when at least a first condition is fulfilled, wherein the first condition includes that the intra prediction mode of the left neighbor block and the intra prediction mode of the above neighbor block are the same angular mode, and wherein ang represents the intra prediction mode of the left neighbor block or the intra prediction mode of the above neighbor block.