- Home >
- Services >
- Access to Knowledge >
- Trend Monitor >
- Type of Threat or Opportunity >
- Trend snippet: How Facial Recognition Technology is used in practice by law enforcement agencies.
Trends in Security Information
The HSD Trendmonitor is designed to provide access to relevant content on various subjects in the safety and security domain, to identify relevant developments and to connect knowledge and organisations. The safety and security domain encompasses a vast number of subjects. Four relevant taxonomies (type of threat or opportunity, victim, source of threat and domain of application) have been constructed in order to visualize all of these subjects. The taxonomies and related category descriptions have been carefully composed according to other taxonomies, European and international standards and our own expertise.
In order to identify safety and security related trends, relevant reports and HSD news articles are continuously scanned, analysed and classified by hand according to the four taxonomies. This results in a wide array of observations, which we call ‘Trend Snippets’. Multiple Trend Snippets combined can provide insights into safety and security trends. The size of the circles shows the relative weight of the topic, the filters can be used to further select the most relevant content for you. If you have an addition, question or remark, drop us a line at info@securitydelta.nl.
visible on larger screens only
Please expand your browser window.
Or enjoy this interactive application on your desktop or laptop.
How Facial Recognition Technology is used in practice by law enforcement agencies.
Police and INTERPOL, for example, the examiners operate autonomously from the investigation teams, and do not have knowledge of the prosecution that requires them to run facial recognition analysis.
To identify an unknown suspect or person of interest, investigators work with probe images and databases. Probe images are images that are part of the law enforcement investigation and which are submitted to a facial recognition system to be compared to a database. Probe images are usually the photos or movies/stills of suspects or persons of interest. To collect these images, investigators (or digital/face experts) either already have an image of the suspect or they extract it from footage of movies/stills. In any case, law enforcement tries to collect the best- quality image to improve the chance of confirming the identity of the person.
Based on the practices followed by the Netherlands Police, the process for using FRT for law enforcement
Step #1: A (possible) crime is reported or suspected. An investigative team, under the supervision of the public prosecutor, is created and requests warrants to collect images relevant to the crime, including images of the suspect(s). If suspects are detected on the images, the team will try to determine their identity. This can be done by human means – through recognition by people who know the suspects, for instance, police officers or witnesses – or by using facial recognition software with a reference database of known people – for instance, suspects and convicts.
Step #2: If a facial recognition search is required, the investigation team will apply for an FRT investigation through the specialized FRT team. This facial examination team runs FRT software to compare the probe image against one or multiple databases. Before doing so, the facial examiners will first judge the quality of the probe image. If suitable for an FRT search, they will enter the probe into the FRT system, allow the system to do the pre-search analysis and may also provide some notable facial landmarks (centre of the eye socket, etc.) to the software. The examiners then set up the FRT software at a setting that is not too narrow, to avoid false negatives, or too wide, to avoid false positives – which would result in a list of candidates too large to be of use.
Step #3: After the search, the facial examiners analyse the list of candidates provided by the software. They run this last operation manually, deploying their expertise to check if one of the candidate images proposed by the system matches the probe image.
Step #4: If the facial examiners make a possible match, only the probe image and the image
of the possible candidate from the reference database are handed to two facial experts. They perform, independently from each other, a full analysis of the probe and the reference image to determine the similarity/dissimilarity between the two faces. This blind peer review is systematically performed before any positive result is communicated to the requesting investigation team. The facial examiners and experts do not know the exact background to the case, to avoid bias as far as possible. The end result is the final consensus conclusion and is reported to the investigation team as an investigative lead.