A browser-based, sensor-agnostic toolkit for quantifying class separability in multispectral remote-sensing data. Built as a complementary deliverable to an MSc thesis at the University of the Aegean.

Quick links

• 🚀 Try the live app — interactive Gradio app on Hugging Face Spaces
• 📂 Source code on GitHub — full implementation, README, license
• 📖 Full documentation — theory, input format, validation rules, six-step workflow

What the tool does

Spectral Separability Explorer accepts any CSV with per-sample band values and a class label, and produces:

• Per-class spectral signatures
• Boxplots and violin plots per band (units-aware: reflectance, height in metres, temperature in °C)
• Jeffries–Matusita distance matrices with discrete 4-bucket colour coding (Poor / Moderate / Good / Excellent)
• Comparative analysis across band subsets (e.g. RGB vs 5 MS vs 7D = RGB + RedEdge + NIR + nDSM + Thermal)
• Ranked separability table for pair-by-pair drill-down
• ZIP export bundle with results CSV plus an HTML interpretation guide

Eight built-in sensor presets cover the most common multispectral and satellite platforms: MicaSense Altum-PT, RedEdge-MX, RedEdge-MX Dual, DJI Phantom 4 Multispectral, DJI Mavic 3 Multispectral, Parrot Sequoia, Sentinel-2 MSI, Landsat 8/9 OLI+TIRS.

For full theory, input format, validation rules, and the six-step workflow, see the GitHub README.

Citation

If you use this tool in academic work:

@software{koroniadis2026spectral,
  author    = {Koroniadis, Nikolaos},
  title     = ,
  year      = {2026},
  publisher = {Hugging Face},
  url       = {https://huggingface.co/spaces/NickKoro21/jm-separability-toolbox},
  note      = {MSc thesis deliverable, University of the Aegean}
}

Author

Nikolaos Koroniadis — MSc Geography and Applied Geoinformatics, University of the Aegean, RSGIS Lab.

Thesis Supervisor: Dr. Christos Vasilakos.

Released under the GNU AGPL v3.0.