What is spectral imaging?

Spectral imaging refers to a group of analytical techniques that collect spectroscopic information and imaging information at the same time. The spectroscopic information tells us about the chemical makeup at the individual points of the image (pixels) allowing a chemical map of the imaged area to be produced.

Types of spectral imaging include hyperspectral imaging and multispectral imaging. The term chemical imaging is also often used.

What is hyperspectral imaging?

Hyperspectral imaging collects spectral information across a wide range of wavelengths (colours) at the same time as recording the spatial information in an image. The greater information provided from these wavelengths enables more detailed chemical information to be mapped, but requires more expensive instrumentation and more complex data processing than the simpler multispectral imaging (see below).

The data collected are stored as a “hypercube”, which contains spatial data in two dimensions and spectral data along a third dimension for each pixel in the image. As with single-point spectroscopy, a variety of chemometric techniques are used to interpret the results.

What is multispectral imaging?

Multispectral imaging collects spectral information from a limited number of wavelengths (colours) whilst simultaneously recording the image. These wavelengths can be customised for a particular application, which leads to cheaper instrumentation and simpler data processing, but without the flexibility of hyperspectral imaging systems.

What is chemical imaging?

Chemical imaging is just another term for spectral imaging. In practice, there are no differences.

What is remote sensing?

Remote sensing is used to refer to spectral imaging (often over a limited range of wavelengths) from a large distance, typically from satellites or aircraft. It can be used to provide information on vegetation, as well as on urban environments.

Spatial dimensions

Spectral imaging can be carried over a range of spatial dimensions, from imaging from satellites with typical pixel size in the 10’s of metres to imaging on the nanometre scale, for instance with Tip-Enhanced Raman Spectroscopy (TERS).

Wavelength ranges

Many types of spectroscopy can used for spectral imaging. Much work with remote sensing (see above) of the earth operates in the visible and short-wave near infrared regions, whilst applications for precision agriculture will be fully within the NIR region (780–2500 nm).

Instrumentation

A wide range of instrumentation is available, allowing scientists to choose the combination of spatial resolution and wavelength range that they need for a particular application.

For applications covering medium distances (acre or low thousands of m2), unmanned aerial vehicles (UAVs, or “drones”) are increasingly being used. Lightweight multispectral and hyperspectral cameras have been developed especially for use on UAVs.