Industrial Image Processing > Food Industry

Industrial Image Processing in the Food Industry

Hyperspectral imaging provides a competitive advantage when it comes to the quality inspection of foods and food packaging.

The food industry is one of the sectors where quality control and food safety are of paramount importance. Products must reach customers intact and in perfect condition and important data, such as the BBD (best-before date), must be correctly printed on the packaging. Defective or damaged packaging must be detected and avoided to ensure that the quality and the shelf-life of foods remains uncompromised and products are free from foreign objects.

Often, quality control in the food industry is preformed manually and products are sampled and sorted out by hand. These monotonous activities impair concentration. Production flaws or leaks may go undetected and items that should have been rejected might end up on grocery store shelves. Moreover, this method does not account for defects that are invisible to the human eye – for example, air that is trapped in the seams of sealed food packaging.

Using industrial image processing to automate the quality control of foods and food packaging is one way to make quality inspection in this industry safer and more efficient.

2D or 3D cameras and dedicated image processing software automatically inspect and analyze products during production.

Image processing systems detect and process a wide variety of parameters, such as:

• Size, volume, geometry, contours, form
• Printed characters (signs, codes, labels)
• Color and color differences
• Completeness
• Differences in chemical composition
• Dimensional accuracy
• Physical presence
• Position
• Surfaces (including reflective surfaces)
• Correct mounting and installation

Image processing systems monitor 100% of all processed products and automatically remove damaged foods or packaging from production. The speed and accuracy of these systems expedite production and improve both safety and quality control.

Figure 1: Hyperspectral imaging extends the application range of grayscale and color cameras. Hyperspectral cameras record over 100 different wavelengths, covering a much broader color spectrum from UV to infrared light. As a result, materials can be clearly identified and classified based on their chemical composition.


Figure 2: Plastic particles in sugar – true-color image of sugar and two different types of plastic (left: sugar mixed with plastic, right: separate substances)


Figure 3: Plastic particles in sugar – screen shot taken by the hyperspectral analysis software: The two different types of plastic are clearly visible in the sugar after a classifier has been applied (right)


Figure 4: Food inclusions in the sealed seam of cheese packaging.

Hyperspectral imaging extends the range of traditional imaging applications

Hyperspectral imaging significantly extends the performance of automated quality control systems by adding a hyperspectral data to the surface view of an inspected object:

• Hyperspectral imaging covers a variety of wave lengths – from ultraviolet to infrared light.
• Hyperspectral imaging is used to detect material properties that are invisible to the human eye.
• It identifies material differences, e. g., the composition of fat and water.
• It precisely detects foreign objects and contaminants.
• It recognizes inclusions in sealed seams.
• It identifies damaged packaging, even on printed packaging film or foil.

Hyperspectral imaging makes hidden food properties visible. It reveals, for example, information about the fat and water content of a piece of meat. Plastics, for example, can be clearly differentiated from organic materials, to precisely detect foreign objects during production.

Even the quality control of bulk foods like salt or sugar can be substantially enhanced using hyperspectral imaging. Hyperspectral imaging uses the absorption signature of the monitored wavelength range (visible light, UV or infrared range) to detect the smallest impurities, even when their color is nearly indistinguishable from the color of the actual product.

Detecting inclusions in sealed seams

Airtight packaging ensures that food does not perish until it ultimately reaches the customer. Any packaging flaws can cause foods to spoil quickly. Until now trapped air or food inclusions in sealed seams went undetected in products that are packaged in printed film or foil.

Hyperspectral imaging sees through sealed seams of film or foil packaging to ensure that damaged or contaminated products are detected and removed during packaging or production. Hyperspectral imaging guarantees that every product is fully inspected and passes quality control.

Automated and contact-free inspection systems offer quality benefits

Camera systems can be directly integrated into the production workflow to ensure seamless processes at early on in the value chain. Visual and manual controls can be replaced by image processing systems to guarantee a comprehensive automated quality control.

The systems are virtually maintenance-free, have a low susceptibility to interference and are available 24 hours a day, 7 days a week. When high quality products are the key to success in your company, cutting-edge image processing technology significantly contributes to your success.

Applications in hygienic areas

Combined with special enclosures and system components industrial imaging systems can also be used in hygienically demanding environments that are often found in the food industry. Adequate technology is used to protect imaging systems from spray water and extreme temperatures to ensure reliable quality control in areas that are generally off-limits to employees. The same applies to process monitoring applications, for example, in the field of proactive maintenance where the maintenance staff needs to receive automated alerts or error messages when process parameters change.

In summary, industrial image processing systems – and hyperspectral imaging in particular – are used to inspect product attributes and characteristics 

• automatically,
• non-destructively
• and comprehensively.

The correct use of these innovative products guarantees added value when compared to conventional visual and manual inspections.

The future of quality control

In addition to quality control, industrial image processing systems enable the automated recording of both product and process data. The data are mainly used for comprehensive quality control in real time and for monitoring production processes to ideally avoid any interferences or malfunctions.

In addition, the recorded data can be used to unlock hidden economic potential.

Combining recorded data with additional parameters enables targeted analyses for:

• Assessing the quality of products from different suppliers.
• Examining the effects that environmental events have on the quality of raw materials and thus, the final products.
• Optimizing formulations or product designs based on recurring trends or patterns.

Image Source: PSI Technics, Inodo, Optikon, AdobeStock, Fotolia, Perception Park, Kai Myller