Hyperspectral Imaging in the Mining/Purveying Industry: A Review
Updated: Jun 16
Since ancient times gold has been the ultimate symbol of wealth and prosperity. Aside from monetary and ornamental use, gold has practical uses in a number of industries including
These revolve around gold’s unique molecular properties, most notably its inertness. Its many uses combined with its relative scarcity make gold a stable commodity, so much so that it has historically been used to back various currencies around the world.
Mining gold can be a simple exercise, but it doesn’t have to be. Because of the potential profits, gold has attracted the brightest minds and most advanced technology, all hoping to strike it rich. Indeed it drove the population to expand west in the united states during the 1800’s during the great gold rush. Eventually, many of these gold reserves were exhausted. When new deposits were identified in Alaska and South America, they spurred gold rushes of their own. Wherever gold was found, people will follow.
And that’s the crux of the problem. If you want to mine gold today you’re likely going to have
a) travel to remote areas to get it
b) bring everything you need with you
That becomes problematic when you’re dealing with purveying techniques. After all, you either need the means to detect gold on site, or you have to transport test material back to a lab to have it analyzed.
The analysis is sometimes quite involved. In this blog we’ll focus on one type of analysis that can be necessary in the purveying of gold, some problems with the method currently being used, and how hyperspectral imaging technology can be a better option.
Sulfur-containing minerals are common in gold-containing ore, and include pyrite and alunite. The presence of these minerals is problematic for the enrichment of gold, and requires an additional processing step. Therefore, the detection of these minerals is important.
Currently, induced polarization exploration techniques are used to identify gold-containing sulfur deposits for float-separation. This process uses electrical decay estimations to approximate sulfur content and requires a process called “ground truth sampling” where bore samples are chemically analyzed to tune the process. Even then, the results are largely imperfect, and provide little insight into the chemical makeup of the ore being studied.
Some mining operations have tried to use drones to perform RGB color sorting to identify the sulfur content of different piles of ore outside of mines in an attempt to avoid the need to transport samples back to a lab for chemical analysis. These attempts were eventually abandoned once it was determined that
1) The color wasn’t indicative of sulfur content
2) The ore was largely inhomogenous in nature
But they weren’t far off with this approach. After all, if a sensor could be mounted to a drone and data could be taken on site without transporting samples that would be an ideal situation. They just weren’t in the right frequency range.
· RGB visible range – Cannot detect sulfur
· NIR range – can detect sulfur
If instead of a RGB sensor they mounted an NIR sensor in the range of 900-1700nm they could detect sulfur. Using hyperspectral imaging, this information could be quantified and mapped. This would allow you to
1) Determine 2D distribution of sulfur
2) Keep the analysis on-site
3) Dispose of current methods
In this way hyperspectral imaging could revolutionize the purveying of gold in remote areas by allowing for a robust and reliable detection method of gold-containing sulfur ore. For these reasons hyperspectral imaging is a promising approach to the mining and purveying of gold.