This project presents a Smart Traffic Monitoring System tailored for urban Pakistan, combining adaptive signal control and enhanced pedestrian safety. The system uses locally available sensor modules (e.g. PIR motion sensors, IR/ultrasonic obstacle sensors, inductive loop detectors) integrated with embedded controllers (such as Arduino or Raspberry Pi) to monitor vehicle flow and queue lengths in real time. Object detection algorithms (e.g. lightweight CNNs like EfficientDet-Lite or Tiny-DSOD) are deployed on edge hardware for recognizing pedestrians-especially children or elderly-waiting to cross. When a crossing is requested via push button or detection, all vehicular signals switch to a special blue-light mode, halting traffic to allow safe passage. Communication among sensors, controllers, and a central dashboard is achieved using WiFi or wired links; periodic local data logging and minimal cloud involvement ensures resilience. Evaluations focus on vehicle wait times, pedestrian crossing safety and duration, system latency, and reliability under varying traffic volumes and weather conditions. The proposed system is feasible in Pakistani urban settings, aiming to improve responsiveness, inclusivity, and efficiency of traffic infrastructure.