LIBS Imaging Data Analysis Workshop

Purpose of this course page is to give you the toolset for working with LIBS imaging datas. As the final result you will make customised elemental and mineral maps from LIBS images.

Course table of contents

1. Data and tools

How to set up your computer ready to process LIBS images and where to get the first LIBS images.

Task 1.1. Install Julia

Set up Julia environment. Examples on this page use Julia Pluto notebook environment, but Jupyter notebooks, vscode-julia, or simple commandline Julia are fine too.

A nice guide to set up Julia can be found on MIT Introduction to Computation Thinking course web page at https://computationalthinking.mit.edu/Fall24/installation/. On unix systems you can install julia via juliaup with the commandline:

$ curl -fsSL https://install.julialang.org | sh

Or on windows commandline:

winget install julia -s msstore

The workshop can be completed without needing to learn much programming or specifics of Julia, but with a little programming a lot more is possible.

Task 1.2. Get some LIBS data

Download a LIBS data image from libs-dataset. This datafile is a netCDF format file containing a LIBS hypercube image for you to play with.

2. Start with images

For the first results, let's make elemental heatmaps. For this first example I will just tell you which wavelength to use for a specific element so you can make your first elemental LIBS images.

Task 2.1.

In Julia call

data =	%load libs data of .nc format

3. A look into a single measurement

Looking at a spectrum. Noise and continuum removal. Peak-finding.

Load and look at a single measurement.

plot(wavelengths, spectrum)

3.1. Remove noise, remove continuum

3.2. Reduce, take only important looking spectral peaks

4. Process the data

In this task we expand the previous task to the full data.

4.1. Preprocess the full data

4.2. Improved elemental maps with reduced data

5. Classify

This task goes back to the big picture. What do we have in our LIBS image here? A rock is composed of different minerals. So the natural think to look for here is what minerals do we have on the rock?

There are many classification methods that work well with LIBS data. Here we use Spectral Angle Mapper

5.1. Find and select materials

5.2. Make a map or two

5.3. Interpretation

Select the map or maps you like.

6. What's in next lessons?

This concludes th