automatic translation

Artificial vision for food quality control

Professionals in the agricultural and packaging sector must respond to a growing demand for high quality food. For this reason, the inspection of food products during the packaging process is essential.

The benefits of accurate quality control

The quality control of food production makes it possible to limit waste, ensure a final product that complies with current regulations, protect the company from any complaints and the withdrawal of non-regulatory products.

However, the challenges are multiple and complex. The inspection must facilitate the identification of products contaminated with foreign bodies, damaged or incorrectly labeled packaging and spoiled food.

Technologies available

To meet these needs, companies rely on technology. More specifically, SWIR (Short Wavelength InfraRed) imaging and X-ray analysis with the use of sources and detectors suitable for use in processing or production lines.
Thanks to video cameras and sensors, it is in fact possible to create automated devices for inspection, sorting and quality control.

Visible and invisible

Artificial vision helps reduce errors in the selection and packaging of food, optimizing the entire process. Products are scanned by high-speed imaging devices, such as video cameras with InGaAs or CMOS technology. The former detect the defects present under the surface of the product, the latter the visible defects.

SWIR hyperspectral imaging, what is it?

As anticipated, there is also an emerging technology to perform this type of monitoring. This is SWIR hyperspectral imaging, which allows both spatial and spectral information to be recorded simultaneously. This technology allows both to detect foreign bodies or defects not perceptible to the naked eye as well as to identify the concentration of ingredients and classify foods.

For example, it is possible to identify the amount of water in fruit and vegetables, which is particularly useful since the moisture content can affect the quality of the products and their conservation.

Types of sensors

SWIR imaging is achieved through the use of InGaAs image sensors, capable of detecting the wavelengths of the near infrared. Such sensors are characterized by high frame rate, low reading noise and high sensitivity.
However, they must be divided into two types: linear and area. The former are suitable for online sorting, thanks to the scanning speed and high sensitivity. The latter are used in various applications, including packaging inspection.

Sources: Controllo e Misura,

Read also: Rental robots for the diffusion of automation
Stay up to date on the latest news from the B2B world, follow us on LinkedIn! 

Articoli correlati

Latest news