hylite

Maps minerals in hyperspectral scans of drillcore and outcrop.

Field
Materials & mineralogy
Method
Classical method
Data it takes
imagery, spectra
License
MIT

hylite covers the processing chain for hyperspectral mineral mapping: correction, hull removal and minimum-wavelength mapping. The hklearn component fits machine learning models on top of those features.

It ships no pretrained weights. Models are fitted to each scan, so results reflect the material in front of the instrument and there is no training set from elsewhere to transfer badly.

That also means every scan needs its own fitting step, and interpretation of the resulting mineral maps stays with the analyst.

Upstream project →

Catalog entry last checked 2026-09-08. All models →

Run hylite on your data

The lab runs this on its own compute, in a container, with the inputs and parameters recorded alongside the result. Initial scoping conversations are free.

Contact the Lab