Adaptive Intra Prediction for Non-Square Video Blocks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image and video coding standards, such as HEVC, are limited in their ability to efficiently process non-square blocks and flexible block sizes due to fixed prediction modes, which affects coding efficiency and complexity.
Innovation Solution
The method involves adaptively selecting a current mode set for intra prediction based on block size, shape, and content, using a variable number of most probable modes (MPMs) and assigning different sets of angular modes to the longer and shorter sides of non-square blocks, and employing specific coding methods like fixed length or variable length coding depending on the number of intra prediction modes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If fixed prediction modes are used in current coding standards, then implementation is simpler, but coding efficiency deteriorates for non-square blocks and flexible block sizes
Solution Approach 1:
The patent implements dynamic adaptation of intra prediction modes by selecting from multiple mode sets (first mode set with 35 modes, second mode set with 65 modes, third mode set with 129 modes) based on block characteristics such as size, shape, and content. This dynamic selection allows the system to optimize prediction accuracy for non-square blocks and flexible block sizes while maintaining simpler mode sets for cases where they are sufficient, thereby improving coding efficiency without unnecessarily increasing complexity.
Solution Approach 2:
The patent changes the parameter of prediction mode selection by introducing multiple mode sets with different numbers of angular modes (35, 65, 129) and selectively applying them based on block characteristics. This parameter change enables the system to adapt the complexity of prediction modes to match the specific requirements of each block, improving overall coding efficiency while avoiding uniform complexity across all blocks.
2Reliability
If more intra prediction modes are used, then prediction accuracy improves, but signaling overhead increases
Solution Approach 1:
The patent segments the complete set of intra prediction modes into multiple mode sets (first mode set with 35 modes, second mode set with 65 modes, third mode set with 129 modes). By segmenting the modes and selectively applying only the appropriate set based on block characteristics, the patent reduces the signaling overhead required to indicate the prediction mode while maintaining high prediction accuracy when needed. This segmentation allows the system to use fewer modes for simple blocks and more modes for complex blocks, optimizing the trade-off between reliability and signaling overhead.
3Productivity
If adaptive mode selection is implemented, then coding efficiency improves, but computational complexity increases
Solution Approach 1:
The patent applies local quality by selecting different prediction mode sets based on local block characteristics such as size, shape (non-square detection), and content properties. Rather than applying a uniform complex mode selection process to all blocks, the system applies adaptive mode selection only where needed (e.g., for non-square blocks or blocks with specific content characteristics), thereby improving coding efficiency for relevant blocks while minimizing the computational complexity increase across the entire video stream.
Data Source
AI summary
A method processes video data by intra prediction in a video coding system. In one implementation, the method receives input data associated with a current block in a current picture, wherein the current block is a non-square block and is coded or to be coded by intra prediction. The method then determines a first set of angular intra prediction modes for the longer side of the current block, wherein a number of the first set of the angular intra prediction modes is greater than a number of a second set of angular intra prediction modes for the shorter side of the current block. The method then determines a current intra prediction mode to derive an intra predictor for the current block, and encodes or decoding the current block according to the current intra prediction mode.