Pakistan 2023 Multi-source Open Building Fusion Dataset

In response to the problems of inconsistent time benchmarks, significant differences in geometric expressions, insufficient attribute integrity, and obvious cross source conflicts in existing open-source building data, this paper constructs a multi-source open-source building fusion dataset for Pakistan in 2023.

This data is based on Google Open Buildings, Microsoft Global ML Building Footprints, and OpenStreetMap building data, supplemented by the 2023 building existence probability grid, building height grid, and road network information. It is produced through the process of "preprocessing timeliness verification matching fusion road constraint adjustment quality control". Organize in the form of a unified building master data, recording target geometric contours, source markers, temporal status, matching relationships, adjustment parameters, and supplementary attributes in a single element.