Multi-vehicle cooperative localization using indirect vehicle-to-vehicle relative pose estimation
Résumé
Vehicle localization (ground vehicles) is a fundamental task for intelligent vehicle systems; this paper deals with the issue of multi-vehicle cooperative localization which can bring performance improvement over traditional single vehicle localization. To tackle the problem of vehicle-to-vehicle (V2V) relative pose estimation that is essential for realizing cooperative localization, an indirect V2V relative pose estimation (InDV2VRPE) method is proposed, which overcomes the disadvantages of direct V2V relative pose estimation methods. As part of this InDV2VRPE method, a new map merging method is described. Cooperative localization is realized using this InDV2VRPE method. Real-data experiments demonstrate that the proposed cooperative localization method can work effectively and improve localization accuracy, especially for heterogeneous vehicle systems.