Vision-based Smart Traffic System
Traffic congestion has been the most critical issues in many countries all over the world, especially those highly developed countries, however most of the existing traffic system in Malaysia are conventional system, which is not efficient enough to cope up with the increase of traffic congestion on the road. This project implements an algorithm to detect the traffic density and emergency vehicles using image processing on a prototype model of a road junction as well as a mobile application to show the traffic condition. The system uses Raspberry Pi 4 Model B as the microcontroller and webcams to capture images of the road condition. This system uses image processing to detect the number of vehicles and the presence of emergency vehicles, then assigns the traffic signal duration dynamically, while the mobile application will show the traffic light state, congestion level and the existence of emergency vehicles. The system successfully achieves the objectives of the project, nonetheless there are limitations such as the processing power of microcontroller and the usage of only one detection method, thus future improvement can be done by replacing a microcontroller with higher processing power and adding more detection methods.