A traffic system refers to a network of roads, intersections, and traffic signals that serve as a control for the flow of automobiles. Traffic systems have primarily contributed to the transportation of urban environments. These systems greatly impact efficiency and safety in traveling. Traffic congestion is simply when the number of vehicles utilized surpasses the threshold capacity of road networks leading to prolonged travel times associated with subsequent consumption of fuel and emission of air pollutants. It can also contribute to accidents and stress for drivers. This paper introduces the novel idea of real-time management of traffic around the front of advanced technology. The system envisioned in this study aims at reducing critical urban traffic congestion by optimizing and scheduling signal timings based on real-time traffic conditions. A deep-learning-based video traffic estimation system is proposed in this paper. Because it can detect and count vehicles in real-time, it can thus calculate traffic density and adjust signal timings accordingly. In this regard, it optimizes the flows of traffic and minimizes congestion points to enhance travel times, reduce fuel consumption, and enhance environmental sustainability. To mitigate this problem, this proposed work suggests a solution to optimize signal timing through traffic signals. This is based on deep learning approaches, including real-time applications. Thus the system is carried out to take traffic data by video cameras and utilize deep learning algorithms in order to correctly identify and count the number of vehicles.