On Traffic Density Estimation with a Boosted SVM Classifier

Traffic density and flow are important inputs for an intelligent transport system (ITS) to better manage traffic congestion. Presently, this is obtained through loop detectors (LD), traffic radars and surveillance cameras. However, installing loop detectors and traffic radars tends to be difficult and costly. Currently, a more popular way of circumventing this is to develop a sort of virtual loop detector (VLD) by using video content understanding technology to simulate behavior of a loop detector and to further estimate the traffic flow from a surveillance camera. But difficulties arise when attempting to obtain a reliable and real-time VLD under changing illumination and weather conditions. In this paper, we study the efficiency of using some informative features and the different combinations of the features in describing the traffic density, and propose a real-time VLD by using a boosted SVM classifier to probabilistically determine the traffic density state. We show through extensive experiments that our proposed real-time VLD achieves an average accuracy at around 95% under various different illumination and weather conditions in daytime.