Scalable Video Coding Block Prediction Using Adjuvant Frames

The patent extracts only the necessary motion information and residual data from the lower spatial layer frame, rather than fully decoding the entire reference frame. This selective extraction approach maintains prediction accuracy by preserving essential motion vectors while avoiding the computational overhead of complete frame decoding, thus resolving the contradiction between prediction quality and decoding time.

The patent segments the prediction process into separate components: motion vector extraction from lower layer, independent reference frame decoding, and selective application of motion compensation. This segmentation allows the system to perform only the necessary computations for prediction without fully decoding unnecessary frame data, reducing computational effort while maintaining prediction accuracy.

The patent performs preliminary extraction of motion information from the lower spatial layer frame before the actual prediction process. By pre-extracting motion vectors and identifying relevant blocks, the system prepares necessary data in advance, avoiding redundant computations during the main prediction phase and thus improving processing speed without sacrificing prediction quality.

The patent introduces an intermediary step where motion information is extracted and processed separately from the main decoding pipeline. This intermediary layer acts as a bridge between the lower spatial layer and the enhanced layer prediction, allowing selective processing of only the essential motion data needed for prediction, thereby improving processing efficiency while maintaining prediction accuracy.

The patent extracts only the essential motion vectors and block information from lower spatial layer frames, rather than fully decoding multiple reference frames. This selective extraction maintains prediction accuracy by preserving the critical motion compensation data while eliminating the computational complexity of complete frame decoding, thus resolving the contradiction between prediction quality and computational complexity.

The patent applies partial action by performing motion compensation only on the specific blocks that require prediction, rather than processing entire frames. By applying motion compensation selectively to individual blocks using extracted motion vectors, the system achieves accurate prediction where needed while avoiding unnecessary computations in other areas, thus reducing overall computational complexity.