Video Processing With Modified Cross-Component Chroma Prediction

The patent introduces a cross-component prediction mechanism that uses luma component information as an intermediary to predict chroma component values. This mediator approach allows the encoder to exploit correlations between luma and chroma components, improving coding efficiency without compromising reconstruction accuracy. The prediction value derived from luma serves as an intermediate step before final chroma reconstruction.

The patent modifies the chroma prediction process by introducing adjustable prediction values and modification parameters. The encoder can adjust prediction strength and apply modifications to prediction values based on local content characteristics, enabling dynamic optimization of coding efficiency while maintaining coding performance across different video scenarios.

The patent segments the chroma prediction process into distinct stages: initial prediction value generation from luma, prediction value modification based on local characteristics, and final reconstruction. This segmentation allows each stage to be optimized independently and enables selective application of complex processing only where needed, balancing accuracy improvements with computational complexity.

The patent applies cross-component prediction selectively rather than uniformly across all chroma blocks. By determining when and where to apply prediction modifications based on content characteristics, the system achieves high prediction accuracy for beneficial regions while avoiding unnecessary processing complexity in regions where simple copying suffices.