Intra-Frame Prediction Mode Selection Using Correlated Reference Blocks

The patent calculates correlation values between the current block and adjacent reference blocks before performing intra prediction mode selection. This preliminary correlation calculation identifies candidate reference blocks that are most likely to have similar prediction characteristics, allowing the system to focus computational resources on a subset of prediction modes rather than exhaustively evaluating all possible modes, thus reducing overall computing power consumption while maintaining prediction accuracy.

The patent applies different processing strategies to different spatial locations by utilizing adjacent reference blocks with calculated correlation values. Instead of uniformly evaluating all prediction modes across the entire block, the system focuses on local regions with high correlation to reference blocks, adapting the prediction mode selection process to the specific local characteristics of each block region, thereby reducing unnecessary computations while preserving accuracy in high-correlation regions.

The patent performs correlation calculations between the current block and adjacent reference blocks as a preliminary step before exhaustively searching all intra prediction modes. This preliminary action filters and ranks potential reference blocks based on correlation strength, enabling the system to prioritize prediction mode evaluations that are most likely to yield optimal results, thereby reducing the total processing time required while maintaining high prediction accuracy.

The patent applies partial action by evaluating only the most relevant prediction modes based on correlation thresholds rather than exhaustively evaluating all possible intra prediction modes. When the correlation between the current block and reference blocks exceeds a predetermined threshold, the system performs a reduced set of prediction mode calculations, sufficient to achieve accurate prediction without the full computational burden of checking every mode, thus reducing processing time while preserving accuracy.

The patent incorporates feedback mechanisms by calculating correlation values between the current block and adjacent reference blocks, using these correlation results to guide and constrain the intra prediction mode selection process. The correlation information serves as feedback that dynamically adjusts which prediction modes are evaluated in detail, ensuring that computational resources are allocated to modes most likely to produce accurate predictions, thus maintaining prediction accuracy while reducing overall computing power consumption.

The patent performs preliminary correlation calculations to identify and prioritize candidate reference blocks before conducting detailed prediction mode analysis. This preliminary action filters out reference blocks with low correlation values, allowing the system to focus computational effort on a reduced set of high-correlation reference blocks and their corresponding prediction modes, thereby maintaining prediction accuracy while significantly reducing the computational search space and power consumption.