While shifting lane on the road, the presence of a car in the blindspot can cause many accidents, since the driver does not always turn his head. Therefore, a blindspot car detection is likely to become an essential part of modern vehicles. We developed a program that detects cars in the particular configuration of blindspot using video data taken from the left or right mirror of a car, using on the one hand edge detection and support vector machine (SVM) learning and on the other hand template matching. This makes this program simple, fast and adaptative thanks to SVM learning. The program only uses basical functions of the Ecole des Mines' Camellia open-source image processing library [1], which is close to Intel's IPL library. Thus the program is easy to adapt to another API; it has already been adapted to an embedded system currently in development at NXP Semiconductors (formerly Philips Semiconductors). The source code was tested using the valgrind code checking tool [3] and was validated on real-world video sequences.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    LaRASideCam: a Fast and Robust Vision-Based Blindspot Detection System


    Contributors:


    Publication date :

    2007-06-01


    Size :

    475662 byte





    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    LaRASideCam: a Fast and Robust Vision-Based Blindspot Detection System

    Blanc, N. / Steux, B. / Hinz, T. et al. | British Library Conference Proceedings | 2007


    Blindspot assist system for a vehicle

    European Patent Office | 2021

    Free access

    Blindspot assist system for a vehicle

    HUZAIFA HARARWALA / RUDRESH LONKAR / ATUL ZANJURNE et al. | European Patent Office | 2024

    Free access

    The Robust and Fast Approach for Vision-Based Shadowy Road Boundary Detection

    Tsai, S.-J. / Sun, T.-Y. / IEEE | British Library Conference Proceedings | 2005