Road safety is an important problem that still afflicts our city roads, even though a lot of work has been done into the direction for reducing the amount of accidents. Because of the majority of accidents are due to wrong human behaviour, the safety solutions aimed to improve the intrinsic security of the driving experience, and consequently of the pedestrians, are not enough. Perception systems, that help the driver or even that give the possibility of autonomous action performed by the vehicle, are the next generation automotive safety systems, that potentially could prevent accidents strongly dependent on human errors. The most common and useful perception task for the automotive application is the detection of obstacles. Such task can be achieved, with different characteristics and results, using different kinds of sensors. In this paper the authors analyze the use of one particular sensor for automotive applications: a stereo vision sensor. Stereo vision has several good features that makes it suitable for a low-cost low-power solution in commercial vehicles. It is relatively cheap and low-power; as any vision system, it provides rich visual information; in addition to singlecamera vision systems, it allows a real-time estimate of the 3D structure of the environment, thus provide more precise localization and measure of the detected objects. Of course stereo vision has also some limitations: as any vision-based system, it is sensitive to environmental conditions, for example it is less effective by night, moreover stereo algorithms are not reliable in presence of low-textured areas. These limitations are due to the algorithms more than to the sensor itself, and it is reasonable to assume that research will propose better algorithms in the future. Considering advantages and limitations, we believe that stereo vision is a good candidate for industrial and commercial automotive applications, if a single low-cost lowpower sensor must be chosen. Also the authors analyze the characteristics of a stereo vision sensor, describe an implemented system for real-time object detection and tracking from a moving vehicle, present results obtained during several realtime tests made on a car provided by Centro Ricerche FIAT, comparing the results of stereo vision with the use of a radar mounted on the same vehicle. Finally, they describe a collection of data sets, that can be used for further comparison with other methods.


    Access

    Access via TIB

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Real-time stereo vision obstacle detection for automotive safety application


    Contributors:
    Perrone, D. (author) / Iocchi, L. (author) / Antonello, P.C. (author)


    Publication date :

    2010


    Size :

    6 Seiten, 8 Bilder, 12 Quellen



    Type of media :

    Conference paper


    Type of material :

    Print


    Language :

    English




    Real-Time On-Board Obstacle Avoidance for UAVs based on Embedded Stereo Vision

    Ruf, Boitumelo / Monka, Sebastian / Kollmann, Matthias et al. | BASE | 2018

    Free access

    Contrast-invariant Obstacle Detection System using Color Stereo Vision

    Cabani, Iyadh / Toulminet, Gwenaelle / Bensrhair, Abdelaziz | IEEE | 2008


    Real-time on-board obstacle avoidance for UAVs based on embedded stereo vision

    Ruf, B. / Monka, S. / Kollmann, Matthias et al. | BASE | 2018

    Free access

    Genetic Approach for Obstacle Detection using Linear Stereo Vision

    Ruichek, Y. / Issa, H. / Postaire, J.-G. et al. | British Library Conference Proceedings | 2000


    A voting stereo matching method for real-time obstacle detection

    Hariti, M. / Ruichek, Y. / Koukam, A. | Tema Archive | 2003