Spatial Area Segmentation for Selective Media Streaming

The patent applies segmentation by dividing the image data into multiple spatial areas (e.g., first area, second area, third area) in addition to temporal segments. Each spatial area is independently encodable and can be selectively transmitted. The playlist structure is also segmented to include separate URL entries for each spatial area segment, enabling fine-grained control over what content is transmitted and received.

The patent introduces a spatial dimension to the existing temporal segmentation by adding area-based division. Instead of only dividing along the time axis, the system now divides along both spatial (area) and temporal (segment) dimensions, creating a two-dimensional segmentation space that enables selective reception of specific area-time combinations.

The patent extracts only the necessary portions of image data for transmission. By identifying regions of interest (such as areas containing detected objects) and extracting only those specific spatial areas for inclusion in the playlist, the system transmits minimal necessary data rather than complete image data, reducing bandwidth consumption while maintaining quality where needed.

The patent applies local quality by allowing different spatial areas to have different transmission priorities and qualities. Regions of interest can be transmitted with higher quality or at all, while non-critical areas may be omitted or transmitted at lower quality, optimizing the overall transmission efficiency based on local importance rather than uniform treatment of all areas.