De La Rue’s Smartphone Authentication App for Holograms Leverages Power of Visual-AI

In last month’s issue of Authentication News®, we looked at the tantalising prospect of combining the analytical capability and connectivity of a smartphone with the authenticating power of a hologram, touching on how reliance on end-user knowledge of hologram features could be alleviated by the use of smartphones to verify the holograms.

Now, De La Rue Authentication has launched DLR Validate™, a smartphone-enabled hologram validation app. The app enables authorised brand protection inspectors to scan the features in a PURE™ (embossed) or IZON® (photopolymer) holographic label to verify if the hologram is genuine. The app is in the form of an optional plug-in to the company’s existing Traceology® product authentication app, and is intended to provide an extra layer of confidence in hologram verification.

Developed in partnership with VISUA, which is headquartered in Dublin, Ireland, DLR Validate leverages VISUA’s enterprise-grade visual-AI* (artificial intelligence) technology, to deliver what is described as unprecedented accuracy at scale, thanks to the patented Adaptive Learning Engine at its core.

The collaboration between the two companies has led to a solution that incorporates image processing and a rule engine (ie. a piece of software that executes rules according to a particular algorithm) to leverage the secure and repeatable nature of genuine PURE or IZON holograms. The result is a tool that can identify counterfeit labels in the market, even those deemed ‘good enough’ to fool the human eye.

With configurable parameters that include image volume, image diversity and batch confidence for high rates of precision (mistaking a fake label for a genuine one) and recall (recognising genuine labels), De La Rue says that DLR Validate can achieve rates of false positives and false negatives of, or very close to, 0%.

*Visual AI (also known as computer vision) is a field of computer science that trains computers to replicate the human vision system. This enables digital devices (like face detectors and QR code scanners) to identify and process objects in images and videos, just like humans do. This means that, instead of using text, users can upload an actual image as the query to a search engine and obtain visually similar matches.