Road Network Conflation Based on Radar Tracks
Ruixin Niu, Pramod K. Varshney · 2014
Abstract : Using ground target radar tracks as building blocks, sophisticated road map estimation and fusion algorithms are developed. A maximum likelihood road estimator, as well as its corresponding Cramer-Rao lower bound, is derived. A track to road association approach is developed to align track segments to existing road segments. If a track is associated to any existing road segment, a track to road fusion algorithm is proposed to fuse the track with the corresponding road segment to improve the road map accuracy; otherwise, a new road is added to the road network. Numerical results are provided to demonstrate the effectiveness of the proposed road estimation and fusion approach.