Intra Prediction Mode Grouping for Adaptive Video Block Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video encoding/decoding technologies face challenges in efficiently determining reference regions and intra prediction modes, particularly for high-resolution and high-quality images, leading to suboptimal compression efficiency.
Innovation Solution
The method and apparatus divide intra prediction modes into MPM candidate groups, allowing for adaptive block division and intra prediction mode derivation based on predefined default modes and multiple MPM candidates, with options for applying offsets and replacing unavailable pixels with available ones.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional intra prediction methods are used without adaptive block division, then the encoding/decoding process is simple, but the prediction accuracy and compression efficiency deteriorate for high-resolution images
Solution Approach 1:
The patent divides the current block into multiple sub-blocks based on block size and shape characteristics. Different sub-blocks are formed according to whether the block is square, horizontal, or vertical, and further divided into first and second sub-blocks with specific size relationships. This segmentation enables adaptive prediction processing that improves accuracy without requiring complete redesign of the prediction system.
Solution Approach 2:
The patent applies different prediction modes and reference regions to different sub-blocks within the current block. Each sub-block can have its own prediction mode selected from MPM candidates or non-MPM candidates, allowing local optimization of prediction quality according to the specific characteristics of each region rather than using a uniform approach for the entire block.
2Productivity
If all intra prediction modes are treated equally, then the processing is simple, but the compression efficiency deteriorates
Solution Approach 1:
The patent segments the intra prediction modes into two distinct groups: MPM (most probable mode) candidates and non-MPM candidates. This segmentation allows the encoder to process different mode types with different strategies, improving compression efficiency by focusing computational resources on the most likely prediction modes while maintaining simplicity through structured classification.
Solution Approach 2:
The patent changes the parameter of prediction mode probability by introducing MPM candidates that are statistically more likely to occur than non-MPM candidates. The system processes these modes with different weighting or priority levels, allowing more efficient compression by anticipating and optimizing for the most probable prediction outcomes.
3Measurement precision
If unavailable pixels in the reference region are handled by simple replacement, then the processing is simple, but the prediction accuracy deteriorates
Solution Approach 1:
The patent performs preliminary identification and classification of unavailable pixels in the reference region before the prediction process. By pre-determining which pixels are unavailable and preparing appropriate handling strategies in advance, the system can accurately handle edge cases without adding complexity during the main prediction computation.
Solution Approach 2:
The patent introduces intermediary processing steps that handle unavailable pixels through intermediate representation or substitution mechanisms. Instead of directly dealing with the complexity of unavailable pixel handling, the system uses intermediate structures or placeholder values that simplify the overall processing while maintaining prediction accuracy.
4Adaptability or versatility
If a single reference region is used for all blocks, then the processing is simple, but the adaptability to different block characteristics deteriorates
Solution Approach 1:
The patent segments the reference region determination process into multiple cases based on block characteristics. Different reference regions are determined for different block types (square, horizontal, vertical) and different sub-block configurations. This segmentation enables high adaptability to various block characteristics while maintaining processing simplicity through case-based logic.
Solution Approach 2:
The patent makes the reference region determination dynamic by adjusting the reference region selection based on the specific characteristics of the current block and its sub-blocks. The system dynamically adapts the reference region to match the spatial and structural properties of the block being processed, improving versatility without requiring complex general-purpose algorithms.
Data Source
AI summary
A video encoding/decoding method and device according to the present invention may: determine a reference area for the intra prediction of the current block; derive the intra prediction mode of the current block; and decode the current block on the basis of the reference area and the intra prediction mode.