Map matching algorithm using belief function theory

Map matching algorithms are used to integrate an initial estimated position with digital road network data for computing the vehicle position on a road map. In this paper, a map matching algorithm based on belief function theory is proposed. This method provides an accurate estimation of vehicle position relative to a digital road map using belief function theory and interval analysis. The core idea of the proposed algorithm is to handle only interval knowledge acquired from sensors and to use the multiple hypothesis technique. This technique proves to be relevant to treat junction roads situations or parallel roads. The results on simulated and real data show the usefulness of the proposed method.