This repository is complimentary to a Datarock applied science blogpost that seeks to demonstrate some fully open source methods for processing, manipulating and analysing hyperspectral data sets with relatively modest computing resources.
This notebok was used to apply geometry lookup tables to raw reflectance data cubes in order to generate a georeferenced hyperspectral mosaic in ENVI grid format.
A notebook used to derive normalised difference vegetation index (NDVI) and a crude vegetation mask from the georeferenced hyperspectral mosaic. A memory efficient native ENVI blockwindow approach was used for computation.
A notebook used to apply a pixel-wise convex hull continuum removal to the hyperspectral data mosaic.
A simplistic white mica feature depth and feature position extraction workflow using quadratic functions fitted to a feature window.
All data presented in this repository are taken from the Dugald Block D HyMap survey supplied by the Geological Survey of Queensland. The GSQ hosts a number of spectral products derived from this and other HyMap™ surveys, which can be downloaded here, however access to the raw spectral data requires temporary AWS links for which the reader is advised to contact GSQ directly.