Title
Document anonymization for border guards and immigration services
Author
Bouma, H.
Pruim, R.H.R.
van Rooijen, A.L.
ten Hove, R.J.M.
van Mil, J.D.
Kromhout, B.
Publication year
2020
Abstract
The current capabilities and capacities of border guards and immigration services can be enhanced using technologies that automate the analysis of travel, identity and breeder documents in order to detect fraud. These technologies can be relevant for countering emerging threats in document and identity verification (e.g., forged documents, impostor fraud, morphed faces) at both manual and highly automated border control points (both in the first and in the second line) and in the issuance process of genuine documents. The travel documents (e.g., passports) and breeder documents (e.g., birth certificates) contain personal information, such as name, date of birth and national number. The personal information must be well protected and a data breach must be avoided at all times. One of the ways to protect the personal data is to minimize the sharing of personal data. Anonymization removes the personal information (e.g., by replacing the personal information by a black bar) and can therefore be used to minimize the sharing of personal data. This paper describes the tool that assists border guards and immigration services for the anonymization of travel and breeder documents. The tool consists of a graphical user interface, document detection, keyword recognition, face detection, number detection, barcode detection and masking of personal data. The results show that only 10 annotated images are needed to reach a keyword detection accuracy of 96% and anonymization sensitivity of 93% of the related personal data
Subject
Defence Research
Defence, Safety and Security
Anonymization
Travel documents
Breeder documents
Border guards
Immigration services
To reference this document use:
http://resolver.tudelft.nl/uuid:d61e3053-8138-4ab8-956c-ce0362b299ba
TNO identifier
880325
Source
Counterterrorism, Crime Fighting, Forensics, and Surveillance Technologies, (Edinburgh) digital event, 21 september 2020
Document type
conference paper