Intra Prediction Mode Derivation for Image Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High-resolution and high-quality image data requires efficient compression techniques to reduce transmission and storage costs, as conventional methods struggle with the increased information load.
Innovation Solution
A video coding method and device that employs intra-prediction by deriving multiple intra-prediction modes for a current block, including a second mode based on the position of a target sample, to improve prediction accuracy and reduce residuals, thereby enhancing coding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If conventional compression techniques are used for high-resolution images, then transmission and storage costs increase, but image quality and resolution cannot be maintained
Solution Approach 1:
The current block is divided into multiple regions based on gradient direction, with different prediction modes applied to different regions. This segmentation allows each region to be predicted using the most appropriate mode, improving overall prediction accuracy while maintaining efficient compression
Solution Approach 2:
The patent dynamically selects prediction modes for different samples within a block based on their position and gradient characteristics. Instead of using a single static mode for the entire block, the system adapts the prediction mode locally, achieving better compression efficiency without increasing complexity significantly
2Device complexity
If a single intra-prediction mode is used for the entire current block, then device complexity is reduced, but prediction accuracy for curved boundaries deteriorates
Solution Approach 1:
Different prediction modes are applied to different spatial locations within the current block based on local gradient characteristics. Samples near curved boundaries use modes better suited for curvature, while flat regions use simpler modes, achieving local optimization without global complexity
Solution Approach 2:
The prediction mode parameter changes dynamically across the block based on gradient direction and sample position. By modifying the mode parameter locally rather than globally, the system achieves higher prediction accuracy for curved boundaries while keeping the overall system manageable
3Measurement precision
If multiple intra-prediction modes are derived based on sample position, then prediction accuracy for curved boundaries is improved, but device complexity increases
Solution Approach 1:
The gradient direction and optimal prediction modes are pre-calculated and stored for different positions within the block. During decoding, the system simply retrieves the appropriate pre-computed mode based on sample position, avoiding complex real-time calculations and reducing decoding complexity
Solution Approach 2:
Reference samples from neighboring blocks are copied and used for prediction when appropriate. This copying approach simplifies the prediction process for certain regions while maintaining accuracy, reducing the computational burden on the decoding device
Data Source
AI summary
According to the present disclosure, an image decoding method performed by a decoding device comprises the steps of: deriving a first intra prediction mode of a current block; deriving neighboring samples including left neighboring samples and upper neighboring samples of the current block; deriving a second intra prediction mode of a target sample on the basis of a location of the target sample and the first intra prediction mode of the current block; and performing prediction for the target sample on the basis of the second intra prediction mode of the target sample.