ImpactVision Can Detect Foreign Objects in Food Using Hyperspectral Imagery

ImpactVision, an AgFunder portfolio company, has launched a new application that uses hyperspectral imagery to detect foreign objects in food.

The key benefits of ImpactVision’s system for foreign object detection are that it is non-invasive, suitable for production-grade speeds, detects non magnetic materials i.e. plastic, paper, cardboard, bones, textiles and fibres, IP65 graded (protected from dust ingress and high pressure water jets), integrates with alarms/rejection systems and a web-based analytics service for remote monitoring can be provided.

Why this system?

Lighter foreign objects, such as plastic, paper, bones, cardboard, textiles and insects can go undetected by visual inspection and by X-ray machines, which typically detect objects denser than the food substance itself.

Current settings

ImpactVision’s hyperspectral system can detect objects as small as 0.08 inches in size at a production-grade speed of 29 in/s with a conveyor belt width of 20 inches. The speed, width, and size of the objects detected can all be customized for each use case. Objects underneath the surface of food products generally go undetected by visual inspections and traditional machine vision systems, however, our hyperspectral software can detect and classify these objects by grouping pixels with similar spectral profiles together.

Next steps

1) Send us samples of your foreign objects for calibration, alongside information about the width and speed of your conveyor belt
2) Imaging campaign takes approximately 3 – 5 days
3) Once we have processed the results, we can arrange a site visit to your warehouse or production facility
4) We can then progress with the installation of one commercial system

Get in touch at abi@impactvi.com with any questions and we look forward to working with you to deliver a safe and secure supply chain!

*ImpactVision is an AgFunder portfolio company. Find out more here.*

Want AgFunder as an investor? Learn more.

Leave a Reply

Your email address will not be published. Required fields are marked *