In 2018, vehicles operating in an autonomous mode have been involved in at least five major accidents. Of these, one involved a pedestrian death due to a lack of timely information available to the detection system. Thermal imaging (TI) could have potentially helped with the time of detection of the pedestrian. This paper will argue how TI can provide supplementary information to existing autonomous detection systems to improve their overall performance. It will outline detection considerations for classifying objects in a thermal image which can later be used in sensor fusion applications. A new labelled dataset of thermal images will also be introduced under snowy, overcast, misty, and nighttime driving conditions. A labeled dataset of a golden retriever and a Doberman will be also be introduced. These datasets are publicly available for download.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Leveraging Thermal Imaging for Autonomous Driving


    Beteiligte:
    Miethig, Ben (Autor:in) / Liu, Ash (Autor:in) / Habibi, Saeid (Autor:in) / Mohrenschildt, Martin v. (Autor:in)


    Erscheinungsdatum :

    01.06.2019


    Format / Umfang :

    953391 byte




    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch




    LEVERAGING UNCERTAINTIES FOR DEEP MULTI-MODAL OBJECT DETECTION IN AUTONOMOUS DRIVING

    Feng, Di / Cao, Yifan / Rosenbaum, Lars et al. | British Library Conference Proceedings | 2020


    Leveraging Uncertainties for Deep Multi-modal Object Detection in Autonomous Driving

    Feng, Di / Cao, Yifan / Rosenbaum, Lars et al. | IEEE | 2020


    THERMAL MONITORING IN AUTONOMOUS-DRIVING VEHICLES

    TELPAZ ARIEL / KAMHI GILA | Europäisches Patentamt | 2017

    Freier Zugriff

    AUTONOMOUS DRIVING SYSTEM, AUTONOMOUS DRIVING METHOD, AND AUTONOMOUS DRIVING PROGRAM

    MISAWA HIDEAKI | Europäisches Patentamt | 2023

    Freier Zugriff