Secondary MPM List Generation via Gradient Template Matching

The patent extracts and transmits only the essential residual information that cannot be predicted by the MPM-based intra prediction, rather than transmitting the entire video block. This extraction approach reduces the data quantity significantly while preserving the necessary information for reconstruction, resolving the contradiction between maintaining original quality and reducing data amount.

The patent changes the parameter representation by using gradient-based template matching to identify similar blocks and deriving MPM lists based on these similarities. This parameter transformation allows efficient encoding of high-resolution and high-frame-rate video by exploiting temporal and spatial redundancies, thereby enhancing video quality while managing data amounts effectively.

The patent performs preliminary gradient calculations and template matching to pre-identify similar blocks and construct MPM lists before the actual encoding process. This preliminary action organizes the reference blocks in advance, making the subsequent encoding process more efficient and manageable, thus enhancing coding efficiency while controlling encoding complexity.

The patent applies gradient-based template matching and secondary MPM list generation selectively based on prediction mode and block characteristics, rather than uniformly for all blocks. This partial application approach improves intra prediction accuracy for beneficial cases while avoiding unnecessary computational complexity in cases where simpler methods suffice, thus resolving the contradiction between accuracy and complexity.