Contents Intro Configure Python Simulation Feature Extraction References Intro I have become sick of copy and pasting feature extraction code around so I’ve updated pytsg to include some basic feature extraction methods. To learn this new interface let’s discuss some common feature extraction methods for hyperspectral data, we are not going to touch on dimensionality reduction techniques like PCA as the interpretation is less obvious than scalar extraction techniques such as band ratios, polynomial fitting and gaussian decomposition.

Contents Intro Data Processing All Colours Simplification Logging accuracy tolerable differences consistency Geological Gamut References I want to continue a thought chain that I had here in the last article on the logging of colour. I was wondering how many colour codes do you need and what does a logged colour of red, green or blue actually represent? This analysis uses all the hylogger data across Australia from the NVCL, this for the most part is diamond core in weathered material and consists of 3734 drillholes and chips.

Good Morning All, I’ve been wondering if the method I’ve been using to convert the hyperspectral colour to RGB was accurate and looked into the conversion of spectra to RGB values a bit further. So to compare the accuracy of our spectra to rgb conversion code I’m going to use the data from the A98437 the 2013 Manyingee Uranium Annual report We will compare the mean colour of the chips from the images to the rgb channel responses as used in the previous post and the correct method using the colour matching functions.

Contents Intro Configure Python Get Data Load Data Merge Data assay validate assay and spectral merge geological data logged colours Assess Geological Data References Intro Let’s run though an example of how to validate unseen spectral data. I was originally going to write about feature extraction but that can be for the next installment.