Applications for Hyperspectral Imaging

Hyperspectral imaging is a field with the potential to service a multitude of applications across many industries. Hyperspectral imaging measures the quantity of light emitted, reflected, and transmitted from an inspected object to generate a spectral signature that includes ultraviolet to near infrared information. Inspection systems use this spectral signature, which is similar to a finger print in its uniqueness, to identify characteristics imperceptible to the human eye. For example, hyperspectral imaging can pinpoint the composition, quality, purity, and age of an object.

In the food industry, one application for hyperspectral imaging is quality assurance. Hyperspectral imaging can identifying discolorations and mold in nuts at a 100% inspection rate, as well as separating out foreign objects such as pieces of shell and bugs. For grading of meat, poultry, and fish, hyperspectral imaging provides insight on protein and lean-fat content and tenderness.

In these applications, hyperspectral sensors help the food processing industry improve product quality, lower the risk of sending unsafe or lower-grade product to market, and reduce inspection costs. The food industry relies heavily on maintaining reputation for quality and safety, and hyperspectral imaging is one way to help lower risk.

Similarly, pharmaceutical companies can’t risk their reputation or the business costs associated with releasing a defective product to market. Hyperspectral imaging paired with machine learning capabilities can distinguish between visually identical but chemically different pills. This could include sorting pills before they are packaged, or detecting subtle changes in the composition of active ingredients in visually identical pills to screen out-of-specification products.

In military applications, hyperspectral imaging can alert a vehicle driver or crew commander of changes in the environment. For example, hyperspectral data can identify camouflage from natural foliage and highlight recently overturned soil to alert on potential roadside improvised explosive devices (IEDs).

Forensics is another prospective field for hyperspectral imaging. Traditional chemical-based methods used in bloodstain and gunpowder residue analysis can be easily contaminated during testing processes. Hyperspectral sensor technology has been proven to locate bloodstains with a higher degree of effectiveness than chemical methods. Forensic scientists have used hyperspectral imaging data to analyze car paint left at the scene of a hit and run to help police identify suspects.

With the ability to accurately see far beyond what is perceived by the human eye or traditional vision systems, hyperspectral imaging will be increasingly adopted across numerous applications. Pleora and perClass BV recently announced a technology partnership to simplify the deployment of machine learning hyperspectral imaging for inspection applications.