Source Cooperative (source.coop)

🤖 AI Summary
A new massive geospatial dataset called Source Cooperative has been released, combining building footprint data from three major sources: Google's V3 Open Buildings, Microsoft's GlobalMLFootprints, and OpenStreetMap. The integrated dataset includes over 2.7 billion building footprints, making it one of the largest and most comprehensive collections available. Each footprint remains tagged with its original source to maintain data provenance and enable fine-grained analysis or source-specific filtering. This collaboration is significant for the AI and geospatial ML community as it offers standardized access to high-quality building outlines at an unprecedented scale, facilitating tasks like urban planning, disaster response, and mapping. The dataset is hosted in cutting-edge, cloud-native geospatial formats—including GeoParquet, FlatGeobuf, and PMTiles—that allow efficient querying, streaming, and scaling for big-data workloads. This will empower researchers and developers to build improved models for object detection, change detection, and spatial analysis, leveraging multi-source data fusion on a planetary scale. By merging these authoritative sources into a single, partitioned dataset with explicit source labels and modern storage formats, Source Cooperative represents a crucial step toward harmonizing global building footprint data. It not only enhances accuracy and coverage but also sets a new standard for interoperability and usability in geospatial AI applications.
Loading comments...
loading comments...