Video Coding Prediction Block Adaptation for Obstacle Interference
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding standards face inefficiencies when dealing with prediction blocks that include obstacles, leading to increased prediction errors and reduced encoding accuracy, particularly in bidirectional prediction modes where one reference block contains an obstacle.
Innovation Solution
A video coding device determines partial regions with high prediction errors in the luminance component and applies a unidirectional prediction mode to corresponding color-difference components, switching between bidirectional and unidirectional prediction modes for each partial region to generate more accurate prediction blocks, thereby reducing the influence of obstacles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a bidirectional prediction mode is used for encoding a block, then coding efficiency is improved by reducing the influence of noise, but prediction accuracy deteriorates when one reference block contains an obstacle
Solution Approach 1:
The current block is divided into a first region and a second region based on motion vector differences. The first region uses bidirectional prediction mode while the second region uses unidirectional prediction mode, allowing different prediction strategies for different parts of the block depending on obstacle presence
Solution Approach 2:
Different prediction modes are applied to different regions within the same block based on local characteristics (motion vector difference thresholds). Regions with large motion vector differences (likely containing obstacles) use unidirectional prediction, while other regions use bidirectional prediction
2Manufacturing precision
If a unidirectional prediction mode is used, then prediction accuracy is maintained when obstacles are present, but coding efficiency is reduced compared to bidirectional prediction
Solution Approach 1:
The block is segmented into regions requiring different prediction modes. Only the second region (with potential obstacles) uses unidirectional prediction, while the first region uses bidirectional prediction to maintain coding efficiency
Solution Approach 2:
The prediction mode parameter is changed dynamically for different regions based on motion vector difference thresholds, optimizing both accuracy and efficiency by using bidirectional prediction where possible and unidirectional prediction only where necessary
3Reliability
If bidirectional prediction is applied to all regions, then noise influence is reduced, but the influence of obstacles in reference blocks increases prediction errors
Solution Approach 1:
The prediction approach is adapted locally based on motion vector characteristics. Regions with large motion vector differences (indicating potential obstacles) use unidirectional prediction to avoid obstacle interference, while regions with small differences use bidirectional prediction for noise resistance
Data Source
AI summary
A region determination circuit determines, for a first block encoded by referring to a first prediction block generated by applying a bidirectional prediction mode for a first component of a pixel value from among blocks in a coding-target picture included in video data, a partial region to which a unidirectional prediction mode is to be applied for a second component on the basis of a difference value for the first component between corresponding pixels belonging to the first prediction block and the first block. A prediction circuit generates a second prediction block for the second component by applying a unidirectional prediction mode to the partial region and a bidirectional prediction mode to a region that is not the partial region. An encoder calculates a prediction error for the second component between corresponding pixels belonging to the first block and the second prediction block and encodes the prediction error.