作者: Matthias Rapp , Markus Hahn , Markus Thom , Jurgen Dickmann , Klaus Dietmayer
DOI: 10.1109/ITSC.2015.77
关键词:
摘要: Automotive localization in urban environment faces natural long-term changes of the surroundings. In this work, a robust Monte-Carlo based is presented. Robustness achieved through stochastic analysis previous observations area interest. The model uses grid-based Markov chain to instantly changes. An extension by Levy process allows statements about reliability and prediction for each cell grid. Experiments with vehicle equipped four short range radars show accuracy performance improvement dynamic environment.