History-Based CCP Models for Efficient Video Block Coding

The patent implements a history table that stores previously determined CCP model parameters and uses this historical information to guide current CCP model selection. This feedback mechanism allows the system to learn from past coding decisions and improve subsequent coding efficiency while maintaining effectiveness through adaptive model selection based on historical performance patterns.

The patent performs preliminary determination of CCP model parameters by maintaining a history table that is updated before or after coding video units. This preliminary action prepares the system with pre-computed model information, enabling faster and more effective coding decisions in subsequent operations without compromising coding effectiveness.

The patent updates the history table of CCP models periodically - specifically before or after coding each video unit - rather than continuously. This periodic update strategy maintains coding effectiveness by ensuring models are current, while avoiding excessive computational complexity by limiting updates to specific intervals rather than every processing step.

The system maintains its own history table internally and performs self-updating of CCP models based on its own coding experience. This self-service approach allows the system to improve its own effectiveness without requiring external intervention or complex external control mechanisms, thereby reducing overall system complexity while maintaining high coding effectiveness.