osml-imagery-toolkit 2.0.0a1

Toolkit to work with imagery collected by satellites and UAVs

The OversightML Imagery Toolkit is a Python package for image processing and photogrammetry on satellite and UAV imagery. It provides sensor models, orthorectification, tiling, dynamic range adjustment, and map tile generation — built on osml-imagery-io for fast, dependency-light access to NITF, GeoTIFF, JPEG 2000, and DTED formats.

This library contains six packages under the aws.osml namespace:

• photogrammetry: sensor models converting between image (x, y) and geodetic (lon, lat, elev) coordinates
• elevation: loading and composing digital elevation models (DTED, GeoTIFF) with multi-source priority, spatial conditions, and geoid offset correction
• metadata: sensor model builders reading metadata through osml-imagery-io
• image_processing: tiling, map tile generation, orthorectification, dynamic range adjustment, and processing chains
• formats: xsdata-generated Python models for SICD and SIDD XML schemas
• features: geospatial feature indexing and property accessors for imagery-derived detections

Installation

pip install osml-imagery-toolkit

No system libraries, no Conda — just pip. Requires Python 3.10+.

Documentation

• API Reference: awslabs.github.io/osml-imagery-toolkit
• Examples: Scripts and notebooks in the examples/ directory

Contributing

We welcome contributions and suggestions. Please see CONTRIBUTING.md for details.