Vehicle Tracking Using Image Alignment and Haar Transform

The large number of rear end collisions due to driver inattention has been identified as a major automotive safety issue. In this paper, we describe a 3-phase vehicle tracking methodology with a single moving camera mounted on the driver’s automobile as input for use in detecting rear vehicles on highways and city streets. In the first phase, vehicle detection, a rear vehicle is detected by symmetrical measurement and Haar transform. A template is created simultaneously which is used for object tracking employing image alignment techniques in the second phase. The template is continuously monitored to handle abrupt changes in surrounding conditions and is updated when necessary in the third phase. We also present a method to compute the relative distances between the rear vehicles and the driver’s car. Initial experiments demonstrate a successful rate over 90%.

• Type: Conference
• Year: 2007
• Where: MVA
• Authors: Lap-Chi Cheung, Yiu Sang Moon
• Keywords: Computer Vision, MVA 2007, Rear End Collisions, Rear Vehicles, Vehicles