Subblock Partition Candidate Selection in Intra Image Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image encoding/decoding methods require performance improvements, particularly in intra-prediction and subblock-wise partitioning to enhance encoding/decoding efficiency and accuracy.
Innovation Solution
A method and apparatus for image encoding/decoding that configures a candidate group of partition types for a current block, determines a partition type based on a candidate index, and derives an intra-prediction mode for subblock-wise intra-prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If subblock-wise intra-prediction is implemented with multiple partition types, then prediction accuracy is improved, but device complexity increases
Solution Approach 1:
The current block is divided into multiple subblocks with different partition types (e.g., horizontal partition, vertical partition, no partition). Each subblock can have its own intra-prediction mode, allowing localized optimization of prediction accuracy while managing complexity through structured segmentation.
Solution Approach 2:
Instead of applying complex subblock-wise prediction to all blocks, the patent selectively applies different partition types to different subblocks based on local image characteristics. This partial application reduces overall complexity while maintaining prediction accuracy where needed.
2Productivity
If multiple candidate partition types are configured for current block, then encoding/decoding performance is improved, but loss of information increases
Solution Approach 1:
The patent uses a candidate index to select from multiple candidate partition types, where the selection is based on feedback from rate-distortion optimization. The encoder evaluates different partition types and selects the one that minimizes prediction error while considering bitrate constraints, thus improving performance without excessive information loss.
Solution Approach 2:
The patent changes partitioning parameters (partition type, subblock size, coding order) to optimize encoding performance. By adjusting these parameters based on image content characteristics, the system improves prediction accuracy while controlling information loss through adaptive parameter selection.
3Productivity
If subblock-wise coding order is adapted, then intra-prediction encoding/decoding efficiency is improved, but manufacturing precision requirements increase
Solution Approach 1:
The patent implements adaptive subblock-wise coding order where the coding sequence of subblocks is dynamically adjusted based on image content and prediction mode. This dynamic adaptation improves encoding efficiency by processing subblocks in an optimal order, while the precision requirements are managed through standardized coding rules.
Data Source
AI summary
The objective of the present invention is to provide a method and an apparatus for intra-prediction. In addition, the objective of the present invention is to provide a method and an apparatus for intra-prediction on a subblock unit basis. Furthermore, the objective of the present invention is to provide a method and an apparatus for determining a division and an encoding order of subblock units.