Spectro-imagery Principle and applications

Hyperspectral imaging or spectro-imaging combines two remote sensing techniques: imaging and spectrometry. Imaging records the spatial distribution of an energy flux resulting from the reflection and/or thermal emission of a medium. Spectrometry measures variations in energy flux in the electromagnetic domain, depending on the chemical composition of the object under observation, plus surface temperature in the emissive domain. Spectro-imaging thus delivers a complex image of dimension 3, two dimensions of which represent spatial information and the third spectral information.

The richness of spectral information requires precise knowledge of the physical phenomena to be observed, and mastery of the means to be used to extract the desired characteristic.

The purpose of this article is to describe the basic concepts required to understand this technique:

• spectral optical properties of observables ;

• radiative transfer in the Earth-atmosphere system to understand the different radiative contributors to the measured signal;

• main spectro-imaging and pre-processing techniques for converting digital counts recorded by the instrument into physical quantities;

• methods for estimating optical properties.

The electromagnetic range considered covers the optical domain from UV (0.4 μm) to thermal infrared (12.0 μm).