The research programme on Recognition of Digital Information and Fingerprinting (HDIeF) was commissioned by the National Coordinator for Security and Counterterrorism (NCTV).
There is a broad range of technologies that enable automated recognition of moving and still photographs and backgrounds and digital data matching. These technologies are potentially useful for investigative and intelligence services and law enforcement agencies, and for private actors that operate in the Public Order and Safety sector. Because of the huge volumes of data exchanged on the internet every day, searching for specific, relevant information can be like looking for a needle in a haystack. The ability to detect digital information quickly and efficiently is an essential tool for effectively combating cybercrime in general and specific offences in particular.
The objective of this research programme is to stimulate existing and new initiatives to help improve these technologies so that they can be applied in the Public Order and Safety sector. At the start of the programme a study was carried out to determine the state of the relevant technologies: ‘hashing’ (comparing identical files), ‘fingerprinting’ (comparing the file content) and ‘watermarking’ (adding an indicator to files to enable recognition). These basic technologies have the following applications:
- Recognition of persons (recognition of behaviour, establishing identity)
- Recognition of objects
- Recognition of texts
- Recognition of (parts of) files (audio, video, photo)
- Recognition of recording equipment used
- Recognition of hidden messages and social network structures
Subjects falling within these categories that are needed by a party in the Public Order and Safety sector and for which the research and business sectors can offer solutions can be expanded into a project within the HDIeF programme. At present, multiple projects are being carried out, including the following:
- Video Search for Public Order and Safety Domain (VOOV): this project concerns participation in an international benchmark for searching video collections. A few photographs are used to search in a collection of video material to find all the videos in which and times at which a specific person, object or location occurs.
- The objective of the Photo Response Non-Uniformity project (PRNU) is to improve the technology used to establish whether a digital photographic or video image was taken with a specific camera, allowing images from different cases to be collated and connected to each other and to the same camera.
- The Optical Character Recognition (OCR) project involves recognition of textual characters in a graphic environment (e.g. a photograph of a number plate at a flash site). The photographed text is first recognised and then converted into a text file, so that other search tools can then be applied.
The HDIeF programme includes close cooperation with the Web Observation Project within the Internet Investigation Network (IRN) of the Gelderland Zuid police. The IRN is a closed ICT infrastructure that affiliated Public Order and Safety sector employees can join and use to monitor the internet anonymously. It also serves as an information sharing network and its development is open source and community driven.