The rapid growth in urban vehicle numbers has led to severe traffic congestion, which current traffic management systems are not equipped to handle. This paper proposes an adaptive traffic signal control system that utilizes advanced computer vision techniques integrated with enhanced security measures for real-time congestion management. By leveraging existing security cameras, real-time traffic data is collected and processed using image classification and object detection algorithms, with secure data transmission ensured through secure socket layers (SSL) encryption. The system dynamically adjusts signal timings based on current traffic conditions and short-term traffic forecasts. Evaluations conducted at a major urban intersection demonstrated significant improvements, including a 25% reduction in average travel times, a 20% decrease in fuel consumption, and an 18% reduction in pollution levels. This approach presents a scalable and secure solution for effective traffic management in modern cities.