SpectralBridge is scientific Python software for preparing and comparing reflectance across airborne, drone and satellite sensor representations. It serves remote-sensing researchers who need traceable intermediate products, restart-safe processing and explicit quality checks.
Three separate workflows process NEON flightlines, supported local drone imagery, and collections of completed products. Spectral libraries and numerical analysis methods support downstream research. Configured sensor-like bands do not establish agreement with independent sensor observations or field fractional cover.
Documentation · Scientific synthesis · Python API · Issues
The distribution is earthlab-spectralbridge; the Python import is
spectralbridge. Python 3.10–3.12 is the tested release range.
At the 19 October 2026 audit, PyPI did not expose this distribution name.
Use the source installation below. Repository HEAD identifies 2.3.0rc1,
which is a candidate identity rather than a published release. Create a virtual
environment, then run:
git clone https://github.com/earthlab/spectralbridge.git
cd spectralbridge
python -m pip install -e .
python examples/first_example.pyThe tiny example inspects an invented spectral table without downloads or credentials. It verifies software use, not scientific accuracy. Continue the beginner path.
| Your inputs | Guide |
|---|---|
| NEON airborne flightlines | NEON processing |
| Supported local drone TIFF/HDF5 | Drone processing |
| Completed flightline products | Bulk analysis |
| Existing polygon spectral library | Library preflight and reporting |
Review output contracts
and QA evidence
before drawing conclusions. Reusable numerical methods require the analytics
extra; their API guide
separates supported numerical contracts from conditional scientific inference.
Cite the software release actually used and relevant scientific methods. Software citation guidance explains the current author-metadata and archive limitations. The repository's CITATION.cff still requires maintainer-approved authorship; the software paper sources are a prepared manuscript, not an accepted publication.
Read contribution guidance, tests and release gates. The license is GPL-3.0-or-later; see LICENSE, NOTICE and HyTools lineage. Historical research is preserved in provenance and archive.
Read the automated AI transparency statement for the maintainer-reported GPT-3.5 history, prompt-log figures and evidence limitations.