Smart tools streamline identity verification at border crossing points
Leveraging advanced technologies, criminals can produce travel documents that have a very authentic look and feel and falsify their contents, such as personal information and travel history to illegally cross borders. To detect these increasingly sophisticated forgeries at border control points, the tools and methods used by the border guards must be developed continually.
Auto-detecting fraud in breeder documents
The issuance of ID and travel documents like passports is the first step in an identity life cycle. These documents are issued by respective authorities based on breeder documents, like birth certificates or other documents, that are used as a basis to proof somebody’s identity. If breeder documents are manipulated, false identities could be established by issuing identity documents or passports with the falsified data. The EU-funded D4FLY project developed tools for automated analysis of breeder documents, helping experts to detect manipulations and prevent issuance of documents containing false identities, which could later on be used to illegally cross borders. “Our goal has been to augment the capabilities of border authorities in countering emerging threats in document and identity verification (e.g. forged documents or morphed faces) across the whole identity life cycle in both automated and manual border crossing points and in the issuance process of genuine documents,” notes project coordinator Armin Reuter. D4FLY investigated new technologies to enhance the document and identity verification process, laying the foundations for smoother border crossing experiences for travellers.
AI technology for document authenticity verification
Exploiting recent developments in AI technology, researchers developed tools that automatically detect anomalies in travel documents. For instance, a newly developed application by D4FLY enables a border guard to analyse security elements in passports. Building on blockchain technology, it is also possible to read and create passport checking history data related to previous checks. Combining AI technology and optical character recognition, the application detects stamps, checks their authenticity, reads their content (such as entry and exit dates) and deduces the holder’s travel pattern. “Our new tool shortens inspection times. Officials no longer need to manually sift through and evaluate the stamps in passports to understand which countries a traveller has visited, the trip duration and the order of visited places,” remarks Reuter.
Smartphone-based identity verification
The second use case involved an easy-to-use smartphone application to verify passenger identity, eliminating the need to get off the bus. The application uses a distributed digital identity system for both storing and retrieving data. “Travellers have to register in advance at special kiosks where their passports are read, and a photo is taken. The registration system sends the data to the smartphones of authorised border guards. When a bus reaches the borders, officers just walk through the bus and use the dedicated smartphone app to verify the passenger identities by comparing the image taken at that moment with the stored image,” explains Reuter. Additional information, such as enrolled passport data, is also displayed in the application.
Biometric corridors for identity verification on the move
For border controls at airports or cruise terminals, where many passengers arrive in waves, D4FLY developed a prototype of a biometric identity verification corridor allowing identity checks on the move. The solution combines biometric technologies, such as 3D facial image recognition, iris verification and body type analysis based on full-body images taken while travellers are walking through the corridor. The information captured in the corridor is compared with the stored reference data. If they match, travellers do not even have to stop. This system based on multimodal biometrics significantly helps cut long lines at airports and ports. Travellers who will be subject to this verification process would need to voluntarily enrol themselves in the prototype D4FLY self-service kiosk. Presentation attack detection technologies were also developed to prevent the enrolment of false identities. D4FLY helped improve the quality of identity verification in border crossings. Project partners collaborated with end users, gathering useful feedback on social aspects to validate the proposed technologies and better meet end user needs.
Keywords
D4FLY, border, document, blockchain, identity verification, biometrics, border crossing points, optical character recognition
