Intra Prediction Mode Signaling With MFM and MPM Indicators
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image encoding/decoding technologies face inefficiencies in intra prediction mode signaling due to diverse block partitioning and limited reference samples for inter color component prediction, leading to increased bit overhead and performance degradation.
Innovation Solution
The method introduces a Most Frequent Mode (MFM) indicator to determine the intra prediction mode of a current block, and constructs a Most Probable Mode (MPM) list excluding the predetermined mode, using specific neighboring blocks for improved encoding/decoding efficiency and inter color component prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the number of intra prediction modes increases to improve prediction accuracy, then the prediction quality is improved, but the bit overhead for signaling the prediction modes increases
Solution Approach 1:
The patent segments the intra prediction modes into two groups: MPM (Most Probable Mode) candidates and non-MPM candidates. The MPM list contains the most frequently occurring modes derived from neighboring blocks, while non-MPM modes are handled separately. This segmentation allows the encoder to signal only the necessary information - a flag indicating whether the mode is in the MPM list and an index if it is - thereby reducing bit overhead while maintaining access to all prediction modes
Solution Approach 2:
The patent changes the parameter representation by using a two-stage signaling approach: first signaling whether the intra prediction mode is included in the MPM list (using 1 bit), and then signaling the specific mode index only when necessary. This parameter change optimizes the bit allocation based on the probability distribution of mode occurrences, reducing overall bit overhead while preserving all 35 intra prediction modes
2Adaptability or versatility
If block partitioning is diversified to improve coding flexibility, then the adaptability to different image regions is improved, but the prediction mode of neighboring blocks becomes more diversified and less predictable
Solution Approach 1:
The patent creates a universal MPM list construction method that works across all block partitioning types (symmetric, asymmetric, vertical, horizontal). The MPM candidates are derived from available neighboring blocks using a standardized approach that adapts to different partitioning configurations. This universal method ensures reliable prediction mode selection regardless of the specific partitioning pattern used in the image
Solution Approach 2:
The patent dynamically adjusts the MPM list construction based on the actual availability of neighboring blocks. Depending on the block partitioning configuration, different neighboring blocks may be available or unavailable, and the MPM list is constructed adaptively using only the available neighbors. This dynamic approach maintains prediction reliability while supporting diverse block partitioning schemes
3Device complexity
If a fixed and limited reference sample is used for inter color component prediction to simplify the process, then the complexity is reduced, but the performance of inter color component prediction degrades
Solution Approach 1:
The patent performs preliminary classification of the current block into two categories: blocks where the luma and chroma blocks have the same position, and blocks where they have different positions. Based on this preliminary classification, different reference sample selection strategies are applied. This preliminary action enables the use of more reference samples and adaptive prediction for blocks that benefit from it, while maintaining simplicity for blocks where the fixed approach suffices
Solution Approach 2:
The patent applies different prediction strategies to different local regions (blocks) based on their specific characteristics. For blocks where luma and chroma positions differ, the patent uses adaptive reference sample selection and linear prediction models tailored to that specific block's geometry. For blocks with aligned positions, the simpler fixed reference sample approach is used. This local quality approach optimizes prediction performance for each block type while managing overall complexity
Data Source
AI summary
There is provided an image encoding/decoding method and apparatus. An image decoding method according to the present invention may comprise reconstructing a most frequent mode (MFM) indicator indicating whether or not an intra prediction mode of a current block is a predetermined intra prediction mode, deriving, on the basis of the MFM indicator, the intra prediction mode of the current block, and generating a prediction block of the current block by performing intra prediction on the current block on the basis of the intra prediction mode.


