Sobel-Based Intra Prediction Directions for SGPM Transform Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding techniques face inefficiencies in spatial and temporal prediction, particularly in deriving transform sets for intra spatial geometric partitioning mode and non-separable primary/secondary transforms, which affect coding efficiency and quality.
Innovation Solution
Implementing Sobel-based intra prediction direction derivation to determine additional transform sets for intra spatial geometric partitioning mode and non-separable primary/secondary transforms, enhancing video coding efficiency and quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional video coding techniques are used for spatial and temporal prediction, then the coding process is simpler, but coding efficiency and quality are insufficient
Solution Approach 1:
The patent implements dynamic transform set selection based on intra prediction mode. Different transform sets are selected according to the specific intra prediction mode used (e.g., angular, planar, DC modes), allowing the system to adapt transforms to the actual content characteristics rather than using a fixed transform set, thereby improving coding efficiency without excessive complexity
Solution Approach 2:
The patent changes the parameter of transform set selection based on intra prediction mode characteristics. By mapping different intra prediction modes to different transform sets (e.g., using non-separable transforms for certain modes and separable transforms for others), the system optimizes the transformation parameters to match the prediction characteristics, enhancing coding quality and efficiency
2Manufacturing precision
If additional transform sets are derived for intra spatial geometric partitioning mode, then video coding quality improves, but processing complexity increases
Solution Approach 1:
The patent performs preliminary derivation of multiple transform sets based on intra prediction mode characteristics before actual encoding/decoding operations. By pre-determining which transform sets to derive based on the intra prediction mode (e.g., deriving non-separable transform sets for certain modes), the system avoids unnecessary computations during the main processing stages, thus improving quality without proportionally increasing complexity
Solution Approach 2:
The patent derives and prepares multiple transform sets in advance (excessive action), but only applies the necessary transforms during actual processing. By preparing transform sets proactively based on prediction mode and only using them when needed, the system improves coding quality while managing processing complexity through selective application rather than full processing of all possible transforms
3Productivity
If transform sets are selected based on intra prediction mode, then coding efficiency improves, but the selection process becomes more complex
Solution Approach 1:
The patent implements self-service transform selection where the intra prediction mode information automatically determines the transform set without requiring external intervention or complex decision-making processes. The mapping between prediction modes and transform sets is predefined and automatic, allowing the system to select appropriate transforms autonomously based on the prediction mode, thus improving efficiency while keeping the selection process manageable
Data Source
AI summary
Example methods, devices, and computer-readable media are described. An example device includes one or more processors configured to determine to use a spatial geometric partition mode (SGPM) for a current block and perform a Sobel-based intra prediction direction derivation. The one or more processors are configured to determine a first intra prediction mode and a second intra prediction mode based on an output of the Sobel-based intra prediction direction The one or more processors are configured to determine a first set of transforms based on the first intra prediction mode and determine a second set of transforms based on the second intra prediction mode. The one or more processors are configured to decode the current block based on the SGPM, and at least one of the first set of transforms or the second set of transforms.