Automotive radar systems have the capability to measure target range, relative velocity, and azimuth angel simultaneously for all detected objects inside the observation area. Developments of present modern driver assistance and safety systems are based on the implementation of powerful radar sensors into each car. A radar network is considered in this paper that consists of 4 distributed near distance radar sensors, which are implemented behind the front bumper, have a maximum target range of 30 m and cover the full azimuth angle area in front of the car. The radar measurement is processed with an extremely high range resolution and accuracy that is needed for position estimation based on a trilateration procedure. This paper describes a signal processing technique, which is implemented in each radar sensor for target detection and parameter estimation. Additionally, an object classification procedure has been developed which is based on a detailed radar echo signal analysis, some feature extractions and target recognition process. In a realistic test environment, it was shown and demonstrated experimentally that the developed target classification system is able to successfully distinguish between the radar echo signals originate from different object classes. This radar sensor capability is a technical challenge and very important for car manufactures if the sensor can distinguish between pedestrian, bicycles and cars for example.
Target classification based on a 24 GHz radar network
2004
5 Seiten, 11 Bilder, 1 Tabelle, 6 Quellen
Conference paper
English
Radar target classification of commercial aircraft
IEEE | 1996
|Radar Target Classification of Commercial Aircraft
Online Contents | 1996
|Automatic target classification for naval radar
Tema Archive | 2004
|14.30 Target Classification based on Near-Distance Radar Sensors
British Library Conference Proceedings | 2004
|