Density based Traffic Monitoring system in Haze and FOG Conditions

The problem of traffic congestion has increased now-a-day's due to the rapid growth of population in major cities. Overwhelming number of vehicles and insufficient roads are the major causes of traffic congestion. This needs new technologies to be adopted, and a better approach for effective traffic management. In the literature, researchers use conventional methods such as IR sensor, wireless sensor, and Fuzzy logic to measure the traffic density. The main limitations of such conventional methods are that they require personal monitoring of the traffic and ineffective to work in foggy weather. The main aim of this work is to develop a real-time adaptive density-based traffic management system that can quantify number of vehicles on roads under foggy weather conditions. The proposed system involves video acquisition, frame extraction, fog removal and vehicle counting. At first, the video is captured by camera and split into number of frames using frame extraction process. The Dark channel prior (DCP) algorithm is used to remove the fog from each frame and the background subtraction method and certain morphological operations are used to count the number of vehicles in real-time. Based on the vehicle count, the system specifies the time required to clear the traffic. This could facilitate ease traffic flow, save time, and even operate in foggy weather conditions, which is an improvement from the conventional timer-based operations of traffic signals.