An adaptive approach for real-time road traffic congestion detection using adaptive background extraction

Traffic congestion is a situation on road networks that occurs as road use increases. When traffic demand increase, the interaction between vehicles slows the speed of the traffic stream and congestion occurs. As demand approaches the capacity of a road, extreme traffic congestion sets in. Current techniques for road-traffic monitoring rely on sensors which have limited capabilities, inflexibility, and are often costly and disruptive to install. The use of video cameras coupled with computer vision techniques offers an attractive alternative to the current sensors. Vision based sensors have the potential to measure a far greater variety of traffic parameters compared to conventional sensors. This work presents an approach for traffic congestion detection based an adaptive background extraction and edge detection techniques using rang filtering. The proposed work uses a special shadow detection algorithm that reduces the chances of misclassification and enhances the segmentation process. An adaptive background extraction technique is used for better object segmentation. In addition, this approach provides real-time statistical information for traffic surveillance on highways such as, the total number of vehicles on the road and the average speed of those vehicles. The proposed system is capable of detecting cars and vans simultaneously.