Intra Prediction Mode Coding with Neighbor-Based MPM Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image encoding and decoding technologies face inefficiencies in handling high-resolution and high-definition images, particularly in intra prediction mode encoding and decoding processes.
Innovation Solution
The method employs intra prediction mode encoding and decoding techniques that utilize neighbor prediction mode information, including angular difference and mode number, to enhance encoding efficiency by selecting most probable modes and optimizing context models and variable length coding tables.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional intra prediction mode encoding is used, then encoding process is simple, but encoding efficiency is low for high-resolution images
Solution Approach 1:
The patent applies preliminary action by pre-defining most probable mode (MPM) candidates based on neighboring block prediction modes before the actual encoding decision. The encoder prepares a list of candidate modes (PM_DIFF_FLAG, PM_DIFF_VALUE, etc.) in advance, allowing the decoder to reconstruct the same candidate list and efficiently select the correct mode without transmitting all possible mode information, thus improving encoding efficiency while managing complexity.
Solution Approach 2:
The patent changes parameters by introducing angular difference (PM_DIFF_VALUE) and mode difference flags (PM_DIFF_FLAG) as new encoding parameters. Instead of directly encoding the prediction mode index, the system encodes the difference between the current block's prediction mode and neighboring blocks' modes, allowing for more compact representation and improved encoding efficiency for high-resolution images.
2Measurement precision
If more prediction mode information is transmitted, then decoding accuracy improves, but bitstream size increases
Solution Approach 1:
The patent extracts only the essential information needed for prediction mode identification by using difference values (PM_DIFF_VALUE) and difference flags (PM_DIFF_FLAG) instead of transmitting complete mode indices. The decoder can reconstruct the actual prediction mode by adding the difference value to the reference neighboring block mode, thus achieving accurate prediction mode recovery with reduced bitstream size.
Solution Approach 2:
The patent applies partial action by selectively encoding only the necessary prediction mode information based on the actual difference from neighboring modes. When the prediction mode matches the neighboring mode (difference = 0), no additional bits are needed. The system transmits minimal information (difference flags and values) only when necessary, reducing overall bitstream size while maintaining decoding accuracy.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An intra prediction method according to the present invention comprises the following steps: deriving neighboring prediction mode information from a left neighboring prediction mode and a top neighboring prediction mode; deriving an intra prediction mode for a target unit to be decoded, using the derived neighboring prediction mode information; and performing intra prediction for the target unit to be decoded, based on the intra prediction mode. According to the present invention, video encoding/decoding efficiency may be improved.