In this paper, we propose an importance sampling-based localization method that approximates the belief to a Gaussian distribution after update for digital map-based localization. It uses the planned route information for constraining the map possibilities and matches the digital map features and higher features detected by the vehicle. This approach does not constrain the vehicle's sensor setup, does not require as much human effort for mapping when compared to High Definition (HD) maps, is relatively more resilient than low level point cloud approaches since it can rely on features that are relative to the city structure, and is self-aware of when it is uncertain between two or more modes, which is useful when fusing with other state estimation methods. The conducted experiments suggest that the method is adequate for a localization pipeline. The code and dataset used were uploaded to an online repository and are open-access.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    A Gaussian approximation of the posterior for digital map-based localization using a particle filter


    Contributors:


    Publication date :

    2021-09-19


    Size :

    1266761 byte




    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Posterior representation with a multi-modal likelihood using the gaussian sum filter for localization in a known map

    Schoenberg, J. R. / Campbell, M. / Miller, I. | British Library Online Contents | 2012


    Gaussian Diffracted Particle Filter

    Ning, X. / Wang, H. / Ning, Y. | British Library Online Contents | 2010


    Gaussian Mixture Approximation by Another Gaussian Mixture for "Blob" Filter Re-Sampling

    Psiaki, Mark / Schoenberg, Jonathan / Miller, Isaac | AIAA | 2010


    Particle Filter Based Localization of Autonomous Vehicle

    Katwe, Supriya / Iyer, Nalini / Khan, Moin et al. | IEEE | 2021


    Gaussian Mixture Approximation by Another Gaussian Mixture for `Blob' Filter Re-Sampling

    Psiaki, M. / Schoenberg, J. / Miller, I. et al. | British Library Conference Proceedings | 2010