Intra Prediction Mode Remapping for Non-Square Video Blocks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies struggle with inefficient intra prediction for non-square blocks, as they lack sufficient modes to effectively utilize the increased number of prediction directions introduced in newer coding standards, leading to suboptimal compression efficiency.
Innovation Solution
The method involves remapping intra prediction modes from a first set designed for square blocks to a second set tailored for non-square blocks, considering the block's aspect ratio and disabling certain modes based on the block's shape, and adjusting mode numbers to accommodate the remapping.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional intra prediction modes designed for square blocks are used for non-square blocks, then the coding structure remains simple, but compression efficiency deteriorates due to insufficient prediction directions
Solution Approach 1:
The patent changes the parameters of intra prediction modes by remapping mode indices and adjusting prediction direction calculations based on block shape. For non-square blocks, the prediction mode index is remapped using a transformation function that considers the block's aspect ratio, and the prediction direction is adjusted by applying a scaling factor to better match the block geometry, thereby improving compression efficiency without requiring a complete redesign of the prediction framework
Solution Approach 2:
The patent introduces asymmetric treatment of prediction modes based on block shape. Different remapping strategies are applied depending on whether the block is wider or taller, with disabled mode sets configured differently for horizontal versus vertical non-square blocks. This asymmetric approach allows the system to adapt prediction directions to the specific orientation of non-square blocks, improving compression efficiency while maintaining manageable complexity through shape-based categorization
2Measurement precision
If the number of intra prediction modes is increased for non-square blocks, then prediction accuracy improves, but mode selection complexity increases
Solution Approach 1:
The patent applies local quality by configuring different sets of disabled prediction modes based on the specific block shape characteristics. For non-square blocks, certain prediction modes are disabled that are less suitable for the block's orientation, while maintaining full mode availability for square blocks. This localized adaptation of mode availability improves prediction accuracy for non-square blocks while controlling complexity by only modifying behavior where needed
Solution Approach 2:
The patent introduces dynamic mode remapping where the effective prediction mode index is calculated based on the block's aspect ratio and orientation. The remapping function dynamically adjusts which prediction directions are applied to which mode indices, allowing the system to adapt the number and orientation of prediction modes to match the specific geometry of each non-square block, thereby improving accuracy while managing complexity through adaptive selection
3Productivity
If prediction directions are adjusted to match non-square block shapes, then coding performance improves, but computational overhead increases
Solution Approach 1:
The patent changes parameters such as prediction direction angles and mode indices through simple mathematical transformations based on block aspect ratio. These parameter adjustments require minimal computation compared to full prediction calculations, enabling the system to adapt prediction directions to non-square block shapes and improve coding performance while incurring low computational overhead
Solution Approach 2:
The patent performs preliminary determination of the block shape category (square, horizontal non-square, vertical non-square) and pre-configures the appropriate disabled mode sets and remapping functions before actual prediction. This preliminary classification and configuration reduces the computational burden during the prediction execution phase, as the system only needs to apply pre-determined remapping rules rather than evaluating all possible prediction modes, thereby improving coding performance while controlling computational overhead
Data Source
AI summary
Aspects of the disclosure provide method and apparatus for video coding. In some examples, an apparatus includes processing circuitry. The processing circuitry decodes prediction information of a first block from a coded video bitstream. The first block is a non-square block and the prediction information for the first block is indicative of a first intra prediction mode in a first set of intra prediction modes for a square block. Then, the processing circuitry remaps the first intra prediction mode to a second intra prediction mode in a second set of intra prediction modes that is used for the non-square block. The second set of intra prediction modes includes at least the second intra prediction mode that is not in the first set of intra prediction modes. Further, the processing circuitry reconstructs at least one sample of the first block according to the second intra prediction mode.