Reconstructed Video Deblocking With Multi-Pass Boundary Filtering

The patent converts the harmful blocking artifacts into a benefit by using the block boundary information to guide targeted filtering operations. The deblocking filter identifies blocks with significant boundary artifacts and applies filtering only to those regions, transforming the problem of visible artifacts into an opportunity for selective quality enhancement while maintaining compression efficiency.

The patent applies different filtering strengths to different regions of the video data. Instead of uniformly filtering all blocks, it evaluates each block boundary and applies strong, weak, or no filtering based on the artifact severity and block characteristics. This local quality approach ensures that filtering is applied precisely where needed to reduce artifacts while preserving detail in regions that don't require processing.

The patent segments the deblocking filtering process into distinct evaluation and execution phases. It first evaluates block boundaries to identify which blocks require filtering and what strength of filtering is needed, then applies the appropriate filter in a second pass. This segmentation allows the system to avoid unnecessary filtering operations and reduces overall computational complexity while maintaining effective artifact reduction.

The patent implements a multi-pass filtering approach where not all blocks are filtered with the same intensity. It uses evaluation criteria to determine which blocks need filtering and applies partial filtering only to those blocks that meet specific thresholds. This selective approach reduces the total number of filtering operations required while still effectively addressing the most problematic blocking artifacts.