Intra-Prediction Video Processing with Selective Sample Filtering

The patent divides the intra prediction process into multiple stages: first intra prediction using reference pixels, then second intra prediction using filtered results. This segmentation allows each stage to focus on specific aspects of prediction, improving overall speed while managing complexity through structured approach.

The patent applies filtering as a preliminary action before the final prediction step. By pre-processing the reference pixels and prediction samples through filtering operations, the system prepares optimized data for the prediction stage, enhancing prediction accuracy and processing efficiency.

The patent merges multiple prediction operations into a unified process where the output of the first intra prediction becomes the input for the second intra prediction. This merging approach processes high-resolution data more efficiently by reusing and refining prediction results rather than treating each prediction independently.

The patent changes parameters such as filter coefficients and prediction modes based on the specific characteristics of the video data and block types. By adapting these parameters dynamically, the system maintains high image quality while optimizing processing requirements for different high-resolution content types.

The patent applies filtering selectively rather than universally. The filter is applied to specific prediction samples based on their position and importance, rather than processing all samples equally. This partial action approach improves prediction accuracy for critical areas while reducing overall processing time.

The patent integrates filtering continuously into the prediction workflow rather than as a separate post-processing step. The filtered prediction samples are used immediately in subsequent prediction operations, maintaining continuous useful action that balances accuracy improvement with processing efficiency.