Hyperspectral Imaging in Industry

Two very popular spectral ranges in industrial spectroscopy are VNIR (400 – 1000 nm) and NIR (900 – 1700 nm). These spectral ranges are appropriate for analyzing, distinguishing and quantifying different materials in respect to their color and molecular structure.

VNIR spectral range

A large part of the VNIR spectral range is associated with the visible part of the spectrum (400 – 800 nm). Consequently, it is not surprising that most of the utilizations of VNIR cameras are related to color measurements. We are surrounded by objects which are colored, and during their fabrication, hyperspectral cameras (which have much better distinctive capability than human eyes) are needed to control their quality.
For instance, do all ceramic floor tiles from the same batch have the same color? Or do died fabrics have the same color from batches to batches? Quality control for color inspection is a key industrial application, and VNIR cameras play a primary role. Hyperspectral cameras can be used for:

• Monitoring of died fabrics, ceramic floor tiles and paint color
• Producing cosmetics or esthetics products: skin color, production of makeup, prosthesis and dying lotions
• Inspecting printing processes, for official documents for instance (bank notes, stamps, visas) or packaging materials; Notice that here, detection of fakes can be done as the same way that real ones are monitored (forensics).
• Displays monitoring (do computer monitors, smart phones, TV or tablets render colors with high fidelity)
• Tattoos removal (the wavelength of the laser used to remove a tattoo depends on the tattoos color)

Notice that in the above applications, the color of the sample or produced good made the quality of the imaged sample or produced goods. Then, measuring its coloration gives a direct and straightforward assessment of its quality. For other applications, the color itself is not the factor of interest, but a mean to monitor a process. For example in cooking. Is a cookie baked too much or is it still raw? Its color would tell you.

VNIR + NIR range

A small portion of the NIR (800 – 1000 nm) is part of the VNIR spectral region, and the combination of visible and NIR spectroscopy makes VNIR cameras a very useful tool also for sorting materials or investigating properties of samples.

In the food industry:

• Fruits inspection: ripeness, bruises, total solid content, water content and firmness can be estimated
• Contaminant detection

In the recycling industry

• PCB
• Plastics sorting (for some of them, but not all kinds, especially if they are dark)
• Sorting of glass and ceramics (for some of them)

In the mining industry

• Sorting of some minerals and ores (especially those related to iron)

NIR spectral range

The VNIR spectral region goes beyond the visible one, and therefore material identification is possible. However, only a tiny part of the NIR is included, and for many industries, this is a limit. The spectral region 900 – 1700 nm encompasses a larger NIR band, and is therefore more prone for sorting materials, especially for recycling purpose.

• For plastics sorting (not black ones)
• For sorting of construction materials (concrete, roof tiles, wood, steel…)

Since the 900 – 1700 nm spectral region is suitable to investigate the molecular composition of samples, it opens the doors to various applications. Among those:

Moisture measurements

• For the production of paper, a thorough monitoring of the water content and moisture level is crucial
• The monitoring of moisture level of combustible (e.g. wood) improves significantly the heat and power efficiency of burners

In the food industry:

• The detection of contaminants (such as fungus)
• The quantification of nutritive properties, such as fat content

In the mining industry

• Many minerals and ores can be sorted

Phenotyping

• Here a combination of VNIR and NIR spectral regions is an asset.

In the pharmaceutics industry

• Quality monitoring, in a qualitative manner: are samples really made of what they should be?
• Quality monitoring, in a quantitative manner: is the composition and concentration of APIs in accordance to specified?

Forensics

• Drugs detection and analysis
• Detection of real and fake products (e.g. pills)

The list of applications presented in this post is open, and many more exists. This list only gives an overview of the possibilities offered by VNIR and NIR spectroscopy, with a focus more oriented towards OEM and industrial usages.