Directional Intra Prediction Using Gradient Weighting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding techniques, such as those in HEVC, while improving intra prediction with directional modes, can be further enhanced to better exploit spatial features and improve performance, especially in regions with high motions or scene changes.
Innovation Solution
The method involves deriving a gradient from neighboring reconstructed pixels along a prediction direction and combining it with non-gradient intra prediction using a weighting factor to form a more accurate intra predictor for pixels in a current block, incorporating vertical, horizontal, slant vertical, and slant horizontal directions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional directional intra prediction modes (DC, Vertical, Horizontal) are used in HEVC, then the coding complexity is reduced and ease of operation is improved, but the prediction accuracy and reliability deteriorate, especially in regions with high motions or scene changes
Solution Approach 1:
The patent introduces gradient calculation as an additional parameter to enhance the prediction process. By computing gradients from neighboring reconstructed pixels and combining them with non-gradient intra prediction through weighting factors, the method enriches the prediction parameters without fundamentally changing the coding structure, thereby improving accuracy while maintaining acceptable complexity
Solution Approach 2:
The patent creates a composite prediction approach by combining gradient-based prediction and non-gradient intra prediction into a unified predictor. This composite method uses weighted summation of both approaches, leveraging the strengths of each to achieve superior prediction accuracy, particularly in high-motion regions where single-method approaches fail
2Reliability
If gradient-based intra prediction is added to HEVC, then the prediction accuracy and reliability improve, but the computational complexity and device complexity increase
Solution Approach 1:
The patent applies gradient-based prediction selectively rather than universally. By using weighting factors to control the contribution of gradient components and applying the method primarily in regions where it provides significant benefit (such as high-motion areas), the approach achieves improved reliability while limiting the increase in computational complexity to only where necessary
Solution Approach 2:
The patent performs gradient calculation and weighting factor determination in advance during the intra prediction process. By pre-computing these values from neighboring reconstructed pixels before the actual prediction, the method prepares the enhanced prediction data structure ahead of time, reducing the computational burden during the main decoding phase and improving overall processing efficiency
Data Source
AI summary
A method and apparatus for intra prediction for a current block using directional intra prediction based on neighboring pixels around the current block are disclosed. Embodiment according to the present invention derives the gradient for the current pixel from the neighboring reconstructed pixels along a prediction direction of directional intra prediction. The gradient derived is used to form the final intra prediction. In another embodiment of the present invention, the intra prediction for slant vertical and horizontal directions is derived based on the neighboring reconstructed pixels in a neighboring row adjacent to the top side of the current block and the neighboring reconstructed pixels in a neighboring column adjacent to the left side of the current block.