Intra Prediction Sub-Partitioning for Image Coding Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image encoding/decoding technologies face challenges in efficiently handling high-resolution images, particularly in intra prediction, where sub-partitioning methods are lacking, leading to suboptimal coding efficiency.
Innovation Solution
A method and apparatus for configuring prediction modes and information for sub-partitioning in intra prediction, allowing for improved prediction signal formation and enhanced coding efficiency by determining prediction modes for target blocks and performing predictions accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If sub-partitioning is applied in intra prediction, then coding efficiency is improved, but device complexity increases due to additional prediction mode determination and partitioning operations
Solution Approach 1:
The target block is divided into multiple sub-partitions (e.g., two sub-blocks) along a division line, allowing independent prediction mode determination for each sub-partition. This segmentation enables more precise local prediction while maintaining overall coding efficiency through selective application to specific block regions.
Solution Approach 2:
Different prediction modes are determined for different sub-partitions based on local characteristics. The first prediction mode for the first sub-partition and the second prediction mode for the second sub-partition are independently selected from available intra prediction modes, allowing local optimization of prediction quality according to regional image features.
2Measurement precision
If multiple prediction modes are determined for sub-partitions, then prediction accuracy is improved, but processing time increases
Solution Approach 1:
By dividing the block into sub-partitions and determining prediction modes independently for each, the patent achieves higher prediction accuracy through localized adaptation while managing processing time through efficient mode selection algorithms and selective application to only those blocks benefiting from sub-partitioning.
Solution Approach 2:
The patent applies sub-partitioning selectively rather than universally - only to blocks where it provides measurable improvement. The processor determines whether to apply sub-partitioning based on block characteristics, performing the additional prediction mode determination only when beneficial, thus balancing accuracy improvement with processing time constraints.
Data Source
AI summary
Disclosed herein are a method, an apparatus, and a storage medium for image encoding/decoding using partitioning. Various prediction methods and prediction modes may be used for prediction of a block. When a block is partitioned, information related to partitioning, such as information indicating whether partitioning is to be performed and information specifying the type of partitioning, is signaled. Also, when partitioning is applied to the block, an available prediction method and an available prediction method in prediction for the block may be restricted in prediction for the block. For optimization of signaling, multiple pieces of information related to prediction may be signaled, wherein the pieces of information may have dependency relationships, and in some cases, additional information may be derived as specific values depending on the value of one piece of information, and signaling of the additional information may be skipped. There can be provided various methods for efficiently signaling multiple pieces of information based on the restrictions and dependency relationships.