Slice-Based Image Encoding with Intra-Slice Merge Control

The patent applies local quality by determining whether to use merge mode on a per-block basis within each slice. The encoder checks if a block can use merge mode by examining whether neighboring blocks are available within the same slice. This selective application allows the system to maintain high encoding efficiency where possible while avoiding cross-slice dependencies that would increase operation time and hinder parallel processing.

The patent changes the parameter of merge mode applicability from a fixed approach to a conditional one. Instead of always attempting to use merge mode, the system changes its behavior based on slice boundaries by checking whether neighboring blocks belong to the same slice. This parameter change allows the system to adaptively apply merge mode only when safe, maintaining parallel processing efficiency while preserving encoding efficiency where possible.

The patent applies segmentation by dividing the image into slices and further into blocks, then controlling merge mode application at the block level within each slice. This segmentation strategy limits cross-slice dependencies by preventing merge mode from accessing motion information across slice boundaries. The system segments the processing task such that each slice can be handled independently, reducing processing complexity while maintaining compression efficiency through localized merge mode application.