Despite the fact that the number of cars on the road is increasing day by day, traffic con- gestion is becoming a serious issue as more vehicles are on the road. A large number of vehicles are waiting to be processed at an intersection due to the increase in traffic flow, and presently the traditional traffic lights are not able to deal with the increasing traffic flow efficiently. As stated above, the theme is to control the traffic by determining the density of traffic on the road and controlling the traffic signals intelligently by using this density information. In order to determine the characteristics of competing traffic flows at signalized road intersections we rely on computer vision and machine learning. In order to achieve this, the traffic control sys- tem uses state-of-the-art, real-time object detection, based on Convolutional Neural Networks called You Only Look Once (YOLO). Finally, traffic signal phases are optimized according to collected data, mainly the queue length and waiting time per vehicle, to enable as many vehicles as possible to pass safely with as little waiting time as possible.