Real Time Image Processing Based Intelligent Traffic Control System

Abstract -Traffic control has become the most serious problem in the large cities. People are running behind schedule in this growing society and wasting time travelling. People need better traffic regulation to have a problem-free, pleasant travel experience and to live in a contamination environment. The project's purpose is to develop a density-based traffic flow signaling system. The signal time adjusts spontaneously when mass transit at the intersection is detected. A web camera is integrated into the traffic system. In order to calculate t, every image acquired by the camera is analyzed. With the rise in flow of traffic, the length of the line of vehicles waiting to be processed at the intersection is rapidly increasing, and traditional traffic lights are unable to adequately plan it. On the highway You Only Look Once, a deep Convolution Neural Network, are being used to compute the amount of vehicles (YOLO).The traffic management system uses Renasas microcontroller which manages the LED's illumination. The proposed system includes RF transmitter and RF receiver. This method is being utilized to detect traffic crashes and automobile rotational violations. When mounted on the road, the entire system helps in peaceful movement of ambulance, fire engine etc., with no necessity for a policeman.

Keywords: Renasas Controller, YOLOv3, RFID, Emergency Vehicle, Accident