Vorausschauende Fahrzeugsensorik mit Photonic Mixer Device und Videokamera für den aktiven Fußgängerschutz (Volume 18) (English shop)

Preview

Table of Contents, Datei (41 KB)
Extract, Datei (98 KB)

Increasing the safety of road users is the focus of much research and development work in the automotive industry. In addition to vehicle occupants, road users also include unprotected road users, that is, among others, all pedestrians. To increase safety, forward-looking sensors for perceiving the vehicle’s surroundings are increasingly being employed. The great challenge in implementing effective safety functions lies both in defining the corresponding function and in selecting the sensor technology and developing the complex sensor data processing. The present work describes an active pedestrian protection system, that is, an active safety function for forward-looking pedestrian protection in motor vehicles. It begins by deriving the requirements placed on the sensor technology and subsequently carries out an analysis of the sensors suitable for use in the vehicle. The task of the sensors and the downstream processing units is the high-performance recognition of pedestrians, that is, detection and classification, on the basis of 3D and 2D image data from moving sensors. The presented system, designated PReSUME (Pedestrian Recognition System Using a Multi-Sensor Environment), achieves high recognition rates while at the same time placing low demands on computing capacity. The selected sensor configuration moreover offers a very high detection rate for all collision-relevant objects with a very low and controllable ghost target rate. The result of the work is thus a sensor system including the real-time-capable image processing software required for the recognition of pedestrians, which, measured against the current state of the art, presents itself as one of the best systems. The work thereby makes a substantial contribution to the implementation of a forward-looking vehicle function for increasing the safety of unprotected road users.

About the author

Björn Elias was born in Leverkusen in 1977 and studied Electrical Engineering and Information Technology at the Rheinisch-Westfälische Technische Hochschule Aachen from 1998 to 2003. Following his diploma thesis, he continued his scientific work at the Chair of Mobile Radio Networks under Prof. Dr. Petri Mähönen. From 2004 to 2007 he was a doctoral candidate at Audi Electronics Venture GmbH in the field of advanced electronics development.

The author has been an employee of AUDI AG since 2007.