Motion Vector Predictor Ordering for Video Coding Efficiency

The patent applies preliminary action by pre-ordering motion vector predictors based on their probability of being the correct predictor. This ordering is established before the actual decoding process, allowing the decoder to quickly select the most probable predictor without complex real-time analysis, thus improving compression efficiency while maintaining relatively simple decoding complexity

The patent replaces traditional complex motion vector prediction mechanisms with a simplified system that uses pre-established ordering and probability-based selection. Instead of using complex spatial or temporal prediction methods, the system substitutes a more efficient predictor ordering mechanism that achieves better compression with less computational overhead

The patent applies segmentation by dividing the set of motion vector predictors into an ordered list where predictors are segmented based on their probability of correctness. This segmentation allows the system to consider multiple predictors for high accuracy while only requiring the decoder to examine them in a predetermined order, thus managing complexity through structured organization

The patent changes the parameter of predictor selection from arbitrary or traditional ordering to probability-based ordering. By changing how predictors are ranked (using probability metrics rather than simple spatial or temporal relationships), the system achieves higher prediction accuracy while the ordering process itself remains a straightforward parameter transformation