Intelligent vision system for obstacle detection based on the three-dimensional high-speed motion analysis

Runtime: 01.05.2009 – 31.07.2011
Sponsor: BMBF – Bundesministerium für Bildung und Forschung (German Ministry of Education and Research) BMBF

Signum Bildtechnik was leading the research work in the frame of HiSpe3D-Vision project, together with the Institute of Automation of the University of Bremen, with the high speed camera producer, the company PCO AG and a well-known Bavarian car manufacturer. The project was focused on the video based driving assistance, aiming to investigate the impact of the high-speed real-time stereo image processing at detection of static and moving obstacles, in order to warn the driver of a vehicle of a danger of an imminent collision. Due to the increased image rate (100Hz), suitable cameras and real-time parallel processing on FPGA, the system prototype is able to detect obstacles which cross the car’s driving pathway up to 50 km/h and estimate collisions with a latency of 0.1s. The efforts are concentrated on the post research tasks towards series implementations, optimizations and marketing.

Contact: Dr.-Ing. Hainfellner (info <a t> signumbt.com)