Unmanned Aerial Vehicles (UAVs) are enabled to be fast and flexible in managing traffic compared to the conventional methods. However, in emergencies, this system takes more time to identify and clear the traffic because of fixed time control. To overcome this problem, an automated intelligent traffic monitoring and controlling system is designed using YOLO V3 neural architecture and implemented to detect the emergency vehicles from video stream data from UAVs using deep Convolution Neural Network (CNN) along with re-routing algorithm to provide the safest alternate route from current position to destination, in a heavy traffic environment. The real-time visual data collected through UAV video cameras are trained using machine learning algorithms to produce statistical profiles that are used continuously as updated inputs to the existing traffic simulation models for improving predictions. The proposed automated system performs exemplary in recognizing emergency vehicles and diverting them to an alternate route for quick transportation in various scenarios.