Fingerprint technique could help prevent banknote fraud
Image credit: Dvmsimages/Dreamstime
Researchers at the University of Warwick have developed a technique that identifies every banknote’s ‘unique and unclonable’ fingerprint, which could help prevent the increasing exchange of counterfeit banknotes.
In 2016, the Bank of England, alongside 50 other countries, introduced plastic (polymer) banknotes to the UK. But as the nation has gradually introduced these polymer banknotes, the number of counterfeit banknotes has also increased.
To prevent this increasing trend, researchers from the Department of Computer Science at the University of Warwick, and their collaborator from Durham University, have proposed a novel technique called Polymer Substrate Fingerprinting, which can identify each banknote’s own unique, unclonable fingerprint.
“Although card and contactless payments are more popular today, banknotes still play a crucial role in society. In fact, there are 500 billion banknotes in circulation in the world, meaning counterfeit notes are a major threat to society and economy,” said Shen Wang, a PhD student from the Department of Computer Science who has been designing, prototyping, and testing this technique for that last two and half years.
The researchers found that every polymer banknote has a unique 'fingerprint', which is caused by the inevitable imperfection in the physical manufacturing process, whereby the opacity coating, a critical step during the production of polymer notes, leaves an uneven coating layer with a random dispersion of impurities in the ink. This imperfection results in random translucent patterns when a polymer banknote is backlit by a light source.
The team has further presented a low-cost method to capture these patterns, by using the commodity negative-film scanner, and processed them into a compact 2048-bit feature vector (fingerprint), to identify each individual banknote with high accuracy. “We have also found the extracted fingerprints contain around 900 bits of entropy, which is dramatically higher than 249-bit entropy for iris-codes used in iris recognition,” Wang added. “This high entropy makes our method extremely scalable to identify every polymer note globally.”
The image analysis focuses on a small feature area, where the random translucent patterns from the opacity coating layer are directly exposed without being obstructed by security printing. For the example of the £10, the chosen feature lies between the ‘ten’ hologram and the see-through window. The random patterns extracted from the opacity coating layer form the unique ‘fingerprint’ is further protected by a veneer finish applied on both sides of the polymer note, and hence, is robust against rough handling in daily use.
Using 340 banknotes the researchers collected an extensive dataset of 6,200 sample images and have proved that their method can identify each banknotes fingerprint successfully despite rough daily handling.
Professor Feng Hao, from the Department of Computer Science, said: “Like every human has unique biometric features, we have found every polymer banknote has its own ‘bio-metric’, which is unique, naturally occurring, and can’t be physically cloned. This new finding gives us the basis to design a completely new anti-counterfeiting method for banknotes.
“Something universally believed is that once counterfeiters have access to essentially the same printing equipment and ink as used by legitimate government to print fake banknotes, the game is over, as there will be no way to distinguish genuine and fake banknotes. Contrary to this belief, our research shows, perhaps surprisingly, that there is still hope to defeat counterfeiting even in that extreme scenario.”
Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.