Tackling smart city traffic congestion

One of the key issues noticed in traffic light points is that many vehicles queue and wait for a long time to cross the traffic light signal even if there is no pedestrian to cross the zebra cross. This time-consuming ineffective traffic control system, as a result, increases the growth of energy loss and harmful gas loss to the environment. In this study, a smart automated traffic control system later named OptiTraffic was developed to capture real-time data and make judgments automatically in order to efficiently and effectively handle the traffic control system. The basic goal of this technology is to automatically detect pedestrians and enable them to cross the road. As a result, reducing the line of vehicles at the traffic light because the traffic signal would not be halted for oncoming vehicles until there are genuine people waiting to cross the road. The smart system utilized deep learning, Raspberry Pi microcontroller, pressure sensors and cameras to collect real-time data and make choices on the pedestrian traffic light management system. The system has cameras along with pressure sensors to validate the pedestrians waiting at the side of the road to cross the road. In this study, the YOLOV5s deep learning model was trained and integrated with the microcontroller to detect pedestrians using the camera. The timer shall turn on for a short period of time when pedestrians get detected at the side of the road. When the stated time of the timer turns on, the system shall turn on the green light for pedestrians in order to allow the pedestrian to cross the road. The input from both the camera and the pressure sensor would act as a confirmation that the pedestrian is certainly willing to cross the road. The final prototype outcome is a complete success as per the previous expectations from the researchers as the real-time detection accuracy of pedestrians was around 88% and also the system was working smoothly overall. The system would overcome the current issues faced by people at the traffic signal points. Lastly, the report concluded with recommendations for further improvements to the developed system.