S. Thoma Fusion von Infrarotvideo- und Laserscanner-Daten zur Fußgängererkennung Diplomarbeit - Technische Universität München (bib)
	This work deals with the detection and tracking of pedestrians from a moving vehicle, especially cars. To estimate the position and moving direction of objects as precise as possible, a laser scanner decive is used in combination with a far infrared camera. We develop methods for efficient segmentation of the 3-dimensional point cloud and optimize them for the given hardware. The well-known multiple hypothesis tracking algorithm is used to ensure optimal object trajectory estimation even in complex scenarios like groups of pedestrians. With Fast Fourier Transform we analyze and detect the pendular movement of pedestrians’ legs. Existing methods for pedestrian detection within far infrared images are integrated. After a feature level fusion of the sensor information, classification is performed using Bayesian networks and decision trees. This paper describes the construction and use of a calibration device which produces a sharp-edged chess board pattern in the infrared image. Thus well-known methods for standard video kameras can be applied in order to determine the intrinsic parameters of the far infrared camera. The implemented techniques are then evaluated with regard to their practical use in road traffic scenarios, by analyzing recorded data from an experimental vehicle. Emphasis is placed on investigating at which distance from the vehicle the methods still work properly.
	This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each authors copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.