Dutch police identified 424 suspects with facial recognition last year
Dutch police have significantly expanded their use of facial recognition technology to identify suspects, relying on a new algorithm that has improved their ability to match images with criminal databases. The system, known as CATCH, facilitated the identification of 424 of suspects in 2023, marking a substantial increase over the previous year.
CATCH, which stands for Central Automated Technology for Recognition, is operated by the Dutch Center for Biometrics in Zoetermeer. It uses a specialized database of facial images collected during criminal investigations, including photos of suspects and convicted individuals.
In 2023, Dutch police submitted 1,693 unidentified images for analysis in criminal investigations. Of these, 1,033 images were deemed suitable for comparison, leading to 424 identifications by CATCH experts. This represents a dramatic improvement from 2022, when only 104 identifications were made out of 913 submitted images.
The police emphasized that strict data protection protocols are followed when using CATCH. “We never employ live facial recognition, and all actions are closely monitored to comply with the Police Data Act,” authorities stated.
Facial recognition through CATCH is primarily used in investigations involving serious crimes, such as sexual offenses, attempted murder, severe assaults, weapons trafficking, and drug-related cases. By comparing images captured by bystanders or surveillance cameras to the CATCH database, investigators can identify suspects who may already have a criminal record.
The technology has proven particularly useful in cases where eyewitnesses or victims provide images of suspects from incidents like street assaults or door-to-door fraud. “Often, suspects have committed crimes before, and this tool helps us connect the dots,” a spokesperson explained.
The Center for Biometrics, which manages CATCH, is also responsible for other biometric identification methods, including fingerprint, handprint, and footprint analyses. Fingerprint identification, known as dactyloscopy, is conducted through HAVANK, a database containing fingerprints from over a million individuals. The system is accessible to other countries and frequently aids in solving cold cases and serious crimes.
Biometric data, defined as measurable physical or behavioral characteristics, include fingerprints, facial scans, and retinal scans. Police officials highlighted the growing role of biometric technologies in modern criminal investigations.
Facial recognition by CATCH is only permitted when identifying a person’s image is deemed crucial to preventing, investigating, or prosecuting crimes. This process may involve either direct comparison between two known images or an automated search in the database when no suspect has been identified.
