Intelligent Traffic Light System Using Image Processing

Abstract

Nowadays, there are more cars used on the road. Traffic congestion problems can cause the economy and the environment both directly and indirectly problems, also part of the problem of air pollution. The traffic light management system in the current situation has used a fixed waiting time, which inability to be flexible according to the traffic at different times such as in rush hour and other. It is not efficient enough to manage traffic with fixed waiting time. The organizers came up with the idea of developing an intelligent traffic light system with flexibility according to the number of cars in real-time by reducing waiting time. The paper was designed and developed by implementing the intelligent traffic light system using image processing technology to process the appropriate waiting time from each image frame. Lazarus and OpenGL were used to program based on Pascal language. The software has been developed for receiving traffic video at the intersection to process car image segmentation of each frame and to calculate the distance of the length of the car in each route in addition. It is also possible to calculate the appropriate time for green-light and red-light duration and corresponding to the length of the waiting vehicles in each route at the intersection. This investigated software can be used to reduce the waiting time at the traffic light intersection by 45.35%. In addition, the intelligent traffic light system is also a social development towards a smart city. The project has created the learning environment and computational thinking for society through the process of STEM Education with using IoT and Artificial Intelligence.