The potential of hyperspectral imaging

Hyperspectral imaging (HSI) is a passive, non-invasive technique that detects reflected light. Combining the high-spectral information from the camera with artificial intelligence software, HSI is used to analyse and detect features in the spatial images. Thanks to its potential, its application domain is growing.

Hyperspectral imaging (HSI) is a passive, non-invasive technology that was originally developed by NASA for remote sensing, satellite and space applications.

Mineral mapping using HSI in visible and near infrared wavelengths.
(Courtesy: USGS - and US DOD)

Other applications include environmental (methane) sensing, forensics, surveillance and so on

HSI has also emerged as a reliable and robust sensor in recycling industry, where it has been used for more than a decade now. Recently HSI is finding increased integration in the food and agriculture industries, which is projected to grow significantly in the next five years.

Quality control of avocado and fat content in meat.

Moisture distribution on a slice of bread and composition of fat, sugar and moisture in sweets
(Source: Campden Bri)


HSI is capable of simultaneously capturing 2D spatial images at multiple wavelengths. In other words, every pixel of a hyperspectral image contains information at the spectral bands that were imaged at.

HSI is a passive, non-invasive technique that fundamentally detects reflected light. Since the reflectance spectrum is a composite of the individual spectra of the constituent materials, the reflectance of an object (or an area of the object) varies depending on the composition. Depending on the application, the spectral range of interest can be ultraviolet (UV), visible to near infrared (VIS-NIR), short-wave infrared (SWIR), medium-wave infrared (MWIR) and long-wave infrared (LWIR). 

Combining the high-spectral information from the camera with artificial intelligence software, HSI is used to analyse and detect features in the spatial images.

Different HSI can decipher different species and other variations in vegetation.
(Source: Chem. Soc. Rev., 2014, 43, 8200)

HSI belongs to a family of spectral imaging techniques. Where high-spectral information may not be a requirement, a variant - multispectral imaging (MSI) - can be a suitable alternative. The main difference between these techniques is that HSI has a large number of contiguous, finer spectral bands, while MSI has a limited number of broad bands of spectrum. To illustrate the differences, let us take the example of vegetation: HSI can decipher the differences in the plant species in a given area of vegetation. On the other hand, MSI can decipher differences between leaves, wood, soil and water. In general, HSI is suitable for an initial study for wavelength selection to design a MSI or in applications where smaller details are a necessity.

Imaging, spectroscopy and imaging spectroscopy
(Ref: Mehta N, Shaik S, Devireddy R, Gartia M. Single-Cell Analysis Using Hyperspectral Imaging Modalities. ASME. J Biomech Eng. 2018;140(2):020802-020802-16. doi:10.1115/1.4038638.)

A quick market search reveals a wide variety in HSI camera offering. Typical questions one may have are: What is the wavelength range suitable for my application? What camera technology best suits my requirements? How do I optimise one of the common technologies such as line-scan, snapshot or tuneable filter for my accuracy, speed and cost requirements?
Sirris can support you in investigating the feasibility of hyperspectral-based vision technology for your application. For any questions please contact us!

Author: Buvana Lefevre