Sistema de detecção de veículos e análise de situação cadastral

Abstract

The use of surveillance cameras is widespread and applied in different situations and environments, being used in large centers including for monitoring the irregular situation of vehicles, however the use of these integrated systems in residential areas requires high investments. The use of theft/stolen vehicles in other crimes has become commonplace, the present work aims to demonstrate the existing capabilities in the field of artificial intelligence, computer vision and machine learning in the detection of vehicles and their registration situation, and these data are used in countless other possibilities. In this work, images captured by cameras installed in the tollbooths of the ES-060 highway, in Espirito Santo, Brazil, under concession of the company Rodovia do Sol (RodoSol), were adopted. Such images underwent treatment and subsequent training of a vehicle detection model and their identification plates. Process that enabled the subsequent recognition of characters and verification of the registration status of the vehicle through ApiPlacas, a system that provides numerous information regarding the vehicles surveyed.