Development of a hybrid vehicle identification system based on video recognition and RFID

The article presents an architecture and experimental implementation of a hybrid vehicle identification system that uses data from a photo-video automatic license plate recognition and an RFID reader. Such a system will significantly increase the likelihood of identifying vehicles that violate traffic rules. In particular, this hybrid system will be capable of recognizing vehicles when license plates are covered with dirt or deliberately hidden. The article presents the system architecture, describes its main components, algorithms for their interaction, and an algorithm for matching data received from the optical recognition system and identifiers received from the RFID reader. We describe an experimental implementation of a hybrid system based on the “Cordon” photo and video recording complex, an experimental RFID reader “P1210” and license plates with tags produced by JSC “Mikron”. In particular, experiments show proper work of the matching algorithm and successful license plate recognition when some characters were changed.