HOW TO RECORD DATA WITH HYPERSPECTRAL CAMERA

This tutorial will introduce how to record data with a hyperspectral camera successfully. We explain the key elements that need to be considered and why. You’ll need a hyperspectral camera, proper illumination, and a reflectance reference that in most cases is a white reference. These are the three most important elements that affect the data.

The spectra you get from your target depend on the camera you use. There are many different hyperspectral cameras with different wavelength ranges.

The acquired spectra show different spectral features if the same target is recorded with two hyperspectral cameras with different wavelength ranges. Different materials and compounds have spectral features in different wavelengths, and one hyperspectral camera won’t bring a solution to every problem.

For example, these three minerals have spectral features in different wavelength ranges. Flat features do not include identifying information, whereas the peaks and shapes do. That is why selecting a camera with an ideal wavelength range for your target is important.

Since hyperspectral cameras measure the light reflected from the target, it is important to use a proper type of illumination. The target should be evenly illuminated, and light should have a spectral response that matches the hyperspectral camera’s wavelength range. The most typical light sources are halogen lights and sunlight. In some cases, LED lights and thermal sources can be used. In conventional digital imaging, the photographer needs to take the color balance into consideration. This is often done using the gray card.

In the same way, the effects of the environment vary in hyperspectral imaging. That is why the recorded data needs to be balanced with the information from the reflectance reference, which is normally a white reference. For example, the white reference can be a panel that is included in the image. Measuring the reflectance reference also corrects the differences in spectral response between hyperspectral cameras. In addition to the reflectance reference, consider the dark reference. It corrects the effects of the camera sensor variations over time. In some hyperspectral camera products, dark reference is collected automatically. Others require user actions.

In the end, we will get the reflectance spectra from the target by correcting data with references with a calculation procedure called reflectance transformation. This calculation makes the data recorded in different environments comparable.