Bone-Dating Using Hyperspectral Imaging!

Headwall recently completed some fascinating demonstration work on behalf of the Conservation Manager and several colleagues at London's Natural History Museum.

One of the hallmarks of hyperspectral imaging is its ability to non-destructively and non-invasively collect an invaluable amount of spatial and spectral data from any sort of reflected matter within the field of view. In commerce and environmental studies, hyperspectral imaging is a valuable and well-known tool that can ‘see’ the unseen.

composite bonesForensics is another exciting area of research. Take 300,000-year-old Neanderthal human bones, for example. Or a 300-year-old snake skin. Or a 400-year-old book of poems. Here you see two bones within the field of view of Headwall's VNIR starter kit. The smaller one is 'only' 200 years old; the larger is 300,000 years old. But the beauty of hyperspectral sensing is that it can classify and compare specimens like these with a tremendous amount of precision, yielding a level of scientific analysis that museums and 'collection-care' experts crave. The demonstration that Headwall performed was an exciting opportunity to show off not only the capabilities of the sensor, but also the capabilities of our new Hyperspec III software. The Conservation Manager was extremely excited with the results of the demonstration. He even remarked that his museum would like to embrace and move forward with the opportunity to be a 'Centre of Excellence for Hyperspectral Imaging,' with Headwall as its sponsor.

Spectral ‘fingerprints’ contain a tremendous amount of useful data, and hyperspectral instruments can see these fingerprints and then extract meaningful data regarding the chemical composition of anything within the field of view. More helpful still, these instruments work in tandem with known spectral libraries that allow a very high degree of selectivity and discrimination. If you know the spectral fingerprint associated with a particular chemical, you can reference it against the hyperspectral data cube coming from the sensor. That fingerprint, once found, very often will be a ‘predictor’ of something else. Disease conditions in crop trees, for example, or the presence of certain inks or pigments on a document or artifact. That’s why precision agriculture and document verification are two other common deployment areas for hyperspectral imaging.