Video Coding Template Matching for IBC and IntraTMP Efficiency

The patent divides the template matching process into multiple stages: initial template matching to generate candidate blocks, followed by refinement using motion information and prediction accuracy evaluation. This segmentation allows the system to achieve high coding efficiency through thorough template matching while managing complexity by processing candidates in stages rather than evaluating all possibilities simultaneously.

The patent introduces motion information as an intermediary element that bridges template matching and prediction accuracy evaluation. Motion information derived from template matching results is used to refine the selection of candidate blocks and evaluate prediction accuracy, enabling the system to achieve improved coding efficiency while maintaining manageable complexity through this intermediate processing layer.

The patent performs preliminary template matching to generate a list of candidate blocks before conducting detailed prediction accuracy evaluation. This preliminary action filters and organizes potential candidates in advance, allowing the subsequent refinement process to focus only on the most promising candidates, thereby improving coding performance while reducing overall processing time compared to exhaustive evaluation.

The patent applies partial action by evaluating prediction accuracy for a subset of candidate blocks rather than all possible templates. By using motion information to guide the selection and evaluation of candidates, the system achieves improved coding performance through thorough evaluation of relevant candidates while avoiding the time cost of evaluating all possible template matches exhaustively.