This paper presents a system for monocular obstacle and lane boundary detection running in real-time. A Kalman filter based depth from motion algorithm is used for the reconstruction of the three-dimensional scene. The proper combination of multiple filters initialized with different states as well as the proposed initialization based on the neighbourhood speed up the rate of convergence. The presented pitch collection stabilises the 3D-reconstruction on typical road scenes significantly. In the authors' current implementation, the 3D-positions of 1100 points are calculated on a 3 GHz Pentium 4 at 15 Hz. This includes the image icquiring process, the tracking of the features on images with QVGA resolution, the Kalman filter estimation using three filters per feature and the obstacle and lane boundary detection. Real world examples illustrate the results of the proposed system.


    Access

    Access via TIB

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Kalman filter based detection of obstacles and lane boundary in monocular image sequences


    Additional title:

    Erkennung von Hindernissen und Fahrbahnberandung in monokularen Bildfolgen auf Kalman-Filter-Basis


    Contributors:


    Publication date :

    2006


    Size :

    7 Seiten, 4 Bilder, 6 Quellen




    Type of media :

    Conference paper


    Type of material :

    Print


    Language :

    English




    Detection of Obstacles in Monocular Image Sequences

    Kasturi, Rangachar / Camps, Octavia | NTRS | 1997


    Detection of Obstacles in Monocular Image Sequences

    R. Kasturi / O. Camps | NTIS | 1997


    Robust lane-markings tracking algorithm based on boundary detection and Kalman filter

    Sun, Tsung-Ying / Huang, Wei-Chih / Tsai, Shang-Jeng | Tema Archive | 2008


    Background Detection in Image Sequences Using the Kalman Filter

    Khilkov, A. V. / Murashov, D. M. / Shamtiev, I. A. | British Library Online Contents | 2003


    Learning of Kalman Filter Parameters for Lane Detection

    Suttorp, T. / Bucher, T. | IEEE | 2006