Detection of railway track crossing violators

An automated system for detecting railway crossing violations, utilizing cameras with facial recognition technology.

The video footage received from the cameras is processed by a neural network to identify any signs of infractions. If a violation is detected, the system compiles an evidence package, which includes a video clip of the incident and photographs of the violator’s face. This evidence is then forwarded to the local Department of Internal Affairs, where authorities identify the violator and issue a resolution on the administrative offense.

Since the project’s inception in the Moscow Region, more than 2,000 individuals have been subjected to administrative penalties, and the number of casualties at railway stations equipped with this surveillance system has been reduced by 90%.