Sistema automático de detecção de infracções rodoviárias com reconhecimento facial do infractor e OCR da matrícula do veículo, na Estrada Circular de Maputo

Abstract

This paper presents the development of a system for detecting and controlling road irregularities, including a facial recognition subsystem for identifying the offender, as well as additionally recognizing the license plate, which is followed by sending the collected information to a server that will make it available to the competent authorities (National Institute of Road Transport) for processing. This system consists of three components, namely: a module for detecting excessive speed, a module for detecting people running red lights, and a module for recognizing faces and license plates. There is a fourth module that coordinates the three subsystems above, interconnecting them via serial communication, and from there, information can be exchanged and decisions made. If a vehicle is moving at a speed above the recommended speed on the road, the speed monitoring module will activate the driver's facial recognition system and the vehicle's license plate recognition system. These same actions are activated when the vehicle is detected passing through a red traffic light. 20 This information (the image of the driver's face and the truth value of its recognition and identification, as well as the vehicle's license plate recognized via OCR (Optical Character Recognition), are sent via e-mail (which can be done by other means) to a server on the Internet (website to be made available by the National Road Transport Institute) where additional processing will take place, including the sending or making available of such information to competent personnel and agencies. (TRADUÇAO NOSSA)