Method for orthorectification of terrestrial radar maps

The vehicle-based Pelican radar system is used in the context of mobile mapping. The R-SLAM algorithm allows simultaneous retrieval of the vehicle trajectory and of the map of the environment. As the purpose of Pelican is to provide a means for gathering spatial information, the impact of distortion caused by the topography is not negligible. This article proposes an orthorectification process to correct panoramic radar images and the consequent R-SLAM trajectory and radar map. The a priori knowledge of the area topography is provided by a digital elevation model. By applying the method to the data obtained from a path with large variations in altitude it is shown that the corrected panoramic radar images are contracted by the orthorectification process. The efficiency of the orthorectification process is assessed firstly by comparing R-SLAM trajectories to a GPS trajectory and secondly by comparing the position of Ground Control Points on the radar map with their GPS position. The RMS positioning error moves from 5.56 m for the raw radar map to 0.75 m for the orthorectified radar map.