Development of Artificial Intelligence Based Vehicle Detection System

Şenol Pazar, Mehmet Bulut, Cihan Uysal
2020 Zenodo  
- Vehicle detection and traffic conditions are important factors for safe driving, accident avoidance, automatic driving and tracking. Currently, Traffic Signal Controller (TSD)systems are used at intersections for traffic signalization operations. Traffic Signal Loop Detectors installed under the road are used in these systems. The installation, operation and maintenance of these detectors are time consuming and costly. Loop detectors can also be damaged if there is any damage on the road. In
more » ... ddition, traffic road information that can be obtained with detectors is also very limited. In this study, which brought a new solution to traffic signalization in many cities in the transportation sector, a vehicle detection system (Video Detector) was developed with an Artificial Intelligence based camera. The main purpose of the study is to develop an image based sensor system as an alternative to conventional under-asphalt magnetic and electrical vehicle sensors. Thanks to the system developed with this study, the need for costly physical loop detectors can be avoided. First phase of this study, vehicle presence and counting can be possible by virtual sensors. Later, vehicle tracking information can be generated depending onvehicle occupation, queuing, average speed, vehicle classification at intersections. Index Terms- Vehicle detection, Traffic signalization, Artificial Intelligence, I ntelligent Transportation Systems
doi:10.5281/zenodo.3922424 fatcat:ywu4vhnw7vfxtn7tybmiwosccu