L-Shaped Intra Prediction Partitions for Lower Video Coding Bit Usage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face inefficiencies in intra prediction, particularly with the increasing number of possible directions, leading to increased bit usage for less likely directions, which hampers compression efficiency.
Innovation Solution
The introduction of non-directional intra prediction modes that partition blocks into L-shaped partitions, utilizing neighboring reconstructed samples for reconstruction, reducing the reliance on specific directional predictions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional directional intra prediction modes are used with increasing number of possible directions, then prediction accuracy is improved, but bit usage increases for less likely directions
Solution Approach 1:
The current block is divided into multiple partitions (e.g., L-shaped partitions), and different prediction modes are applied to different partitions. This segmentation allows the encoder to use directional prediction only where necessary while using non-directional prediction elsewhere, reducing overall bit usage while maintaining prediction accuracy in critical areas.
Solution Approach 2:
The patent introduces non-directional intra prediction modes (such as DC prediction and planar prediction) as alternative parameters to the traditional directional prediction modes. These modes provide a different approach to prediction that does not require specifying a direction, thereby reducing bit usage for modes that are less likely to occur.
2Loss of information
If non-directional intra prediction modes are used, then bit usage is reduced, but prediction accuracy may be compromised
Solution Approach 1:
The patent dynamically selects between directional and non-directional prediction modes based on the characteristics of each partition and the available reference samples. The encoder evaluates which mode is more suitable for each partition and makes adaptive decisions, ensuring prediction accuracy is maintained while minimizing bit usage.
Solution Approach 2:
By dividing the block into partitions, the patent allows different prediction strategies to be applied to different regions. L-shaped partitions can use non-directional prediction when appropriate, while other partitions can use directional prediction, thus maintaining overall prediction accuracy while reducing total bit usage.
3Productivity
If L-shaped partitions are used for non-directional prediction, then compression efficiency is improved, but device complexity increases
Solution Approach 1:
The patent uses L-shaped partitions which are a specific type of segmentation that aligns with the natural boundaries of many video content structures. This segmentation approach improves compression efficiency by better matching the partition boundaries to content characteristics while maintaining relatively simple partitioning logic that can be efficiently implemented in hardware or software.
Data Source
AI summary
Aspects of the disclosure include methods, apparatuses, and non-transitory computer-readable storage mediums for video encoding/decoding. An apparatus includes processing circuitry that decodes prediction information of a current block in a current picture that is a part of a coded video bitstream. The prediction information indicates a non-directional intra prediction mode for the current block. The processing circuitry partitions the current block into a plurality of partitions. The plurality of partitions includes at least one L-shaped partition. The processing circuitry reconstructs one of the plurality of partitions based on at least one of (i) neighboring reconstructed samples of the one of the plurality of partitions or (ii) neighboring reconstructed samples of the current block.