One technique used to analyze artwork is take pictures at various light wavelengths to detect painting alterations and see if there are charcoal drawings underneath the paint pigments.  If a painting has been touched up or altered over the years, pigments that might look the same in normal light might be detected if the camera uses a particular frequency of light.  Similarly, sometimes pigments will become translucent at certain wavelengths allowing you to see something else underneath such as the artist's original sketch.  A hyperspectral camera allows the user to acquire a large dataset with information about the object of interest at many different frequencies.  The dataset can be used to extract many different images of the object at many different frequencies of light.

We used our hyperspectral camera to capture a dataset on part of a painting pictured below.

550nm

610nm

680nm

830nm.  At this frequency you can see the artists original charcoal sketch of dress.

 The MaxMax hyperspectral software will

• Load a series of images taken with our hyperspectral camera
• Automatically register and align the images to create a 3D datacube
• Extract out specified frequency columns for each data slice and assemble them together to form a new image

Using software like ImageJ, you can load the series of frequency extracts into a slideshow allowing you to quickly scan through a large volume of data.