A Novel Method for Effective Vehicle Detection and Tracking Eliminating Occlusion Khilar Rashmita*, Chitrakala S.** *Research Scholar, Department of Computer Science and Engineering, Anna University, Chennai, India **Associate Professor, Department of Computer Science and Engineering, Anna University, Chennai, India Online published on 15 September, 2016. Abstract Intelligent Transportation System (ITS) plays a significant role in constructing a modern transportation to safeguard the lives of people along the busy roads. In ITS, Visual traffic surveillance using computer vision techniques makes it more sophisticated to achieve better traffic control. This paper proposes a new traffic surveillance system to detect and track the vehicles effectively using the video captured by vertical positioned camera. Vehicle detection is based on motion detection, kernel density estimation in a pixel-based technique is used for motion detection. The 3D pose estimation using an Evolutionary computing framework is used to handle all kinds of occlusion. Then the 3D model based tracking is achieved by using Unscented Kalman Filter. The proposed model performs better in foreground object detection, tracking and vehicle speed estimation. Top Keywords Intelligent Transportation system, Vehicle Detection and Tracking, pixel based method, kernel density estimation, Evolutionary computing framework, Unscented Kalman Filter. Top |