Short Range Radar Signal Processing for Lateral Collision Warning in Road Traffic Scenarios
Dirk Bank · IEEE Intelligent Vehicles Symposium · 2007
This paper introduces signal processing algorithms of single-sensor-multi-target-tracking, sensor data fusion, and multi-sensor-multi-target-tracking developed for designing a novel lateral collision warning function. In order to improve the perception of road vehicles, an experimental vehicle has been equipped with arrays of three short range radar sensors on both sides. With the aid of such sensor arrangements, lateral objects (e.g. cars, trucks, bicycles, guard rails) can be detected and tracked. Thus, imminent collisions in the lateral area may be prevented by warning the driver when the ego-vehicle inadvertently drifts towards another vehicle on the adjacent lane or towards the road boundary, or when another vehicle approaches the ego-vehicle in a dangerous manner.