Neighbor-Aware Intra Prediction for Lower-Bitrate Image Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing demand for high-resolution and high-quality images leads to a surge in transmission and storage costs due to higher bit rates, necessitating improved image encoding/decoding efficiency.
Innovation Solution
An image encoding/decoding method that generates a final prediction block by combining intra prediction modes of a current block with those of neighboring blocks, using weighted sums and applying specific weights based on block coordinates and template matching costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If high-resolution and high-quality images are transmitted and stored, then image quality is improved, but transmission costs and storage costs increase due to higher bit rates
Solution Approach 1:
The patent applies position-dependent prediction combination (PDPC) that dynamically adjusts prediction parameters based on spatial position and neighboring block modes. This changes the prediction parameters adaptively to achieve better compression efficiency, reducing bit rate while maintaining image quality.
Solution Approach 2:
The patent uses template matching to evaluate prediction accuracy and provides feedback for selecting optimal prediction modes. The template matching cost serves as feedback to determine whether to apply PDPC and how to combine predictions, improving compression efficiency without sacrificing quality.
2Device complexity
If conventional intra prediction is used for current block, then encoding simplicity is maintained, but prediction accuracy is insufficient leading to higher bit rates
Solution Approach 1:
The patent introduces template matching as an intermediary evaluation mechanism between the current block and neighboring blocks. This intermediary process assesses prediction accuracy and guides the selection of prediction combination strategies, improving prediction accuracy while managing complexity through selective application.
Solution Approach 2:
The patent implements dynamic prediction mode selection where the encoding approach changes based on template matching results and neighboring block characteristics. The system dynamically adjusts between conventional intra prediction and PDPC based on actual prediction needs, balancing complexity and accuracy.
3Device complexity
If prediction block is generated using only current block intra prediction mode, then encoding complexity is low, but prediction precision is insufficient
Solution Approach 1:
The patent merges multiple prediction blocks generated from different neighboring blocks using position-dependent prediction combination. This combining strategy weights and integrates predictions from above and left neighboring blocks with the current block's prediction, improving precision while maintaining reasonable encoding complexity through selective merging.
Solution Approach 2:
The patent applies different prediction combination strategies to different spatial positions within the current block. The PDPC method uses position-dependent weighting where samples closer to neighboring blocks receive different weights than those farther away, achieving local optimization of prediction precision.
4Measurement precision
If PDPC is applied to all blocks, then prediction accuracy is improved, but encoding complexity and computation time increase
Solution Approach 1:
The patent applies PDPC selectively rather than universally - only to blocks where template matching indicates it would be beneficial. This partial application approach achieves prediction accuracy improvement where needed while avoiding unnecessary computation in blocks where conventional prediction suffices, reducing overall encoding complexity and time.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An image encoding/decoding method and device are provided. An image decoding method according to the present disclosure may comprise the steps of: generating a first prediction block of a current block on the basis of an intra prediction mode of the current block; generating a second prediction block of the current block on the basis of an intra prediction mode of a neighboring block of the current block; and generating a final prediction block of the current block on the basis of the first prediction block and the second prediction block.