Traffic clogging emerges as a critical issue due to the burgeoning population and increased vehicular presence in urban areas. The resulting traffic jams not only lead to added stress and delays for drivers but also contribute to elevated fuel consumption and air pollution. While this problem seems pervasive, it disproportionately affects megacities. The relentless growth of this issue necessitates real-time assessment of road traffic density for more effective signal control and traffic management (Discover Artificial Intelligence Journal, 3(1), 39). The traffic controller plays a pivotal role in determining the efficiency of traffic flow, accentuating the need to optimize control mechanisms to cope with the escalating demand. Our proposed system tackles this challenge by utilizing live camera footage from traffic junctions to compute traffic weightage by way of Computer Vision Techniques and Deep Learning. The prime emphasis is to devise an algorithm that revises traffic light patterns based on traffic weight, with the ultimate goal of mitigating choking, improving transit speed and reducing pollution. To overcome the problem, after implementing our proposed system, we observed that it improves overall performance by about 23%, with notably higher efficiency in scenarios with moderately or highly skewed traffic distribution, compared to the current system with fixed signal timings.