Maritime Environmental Perception with FMCW-LiDARs

For autonomous or highly automated ships, it is important to have an accurate picture of all static and dynamic obstacles in a wide area in order to have enough time and space to avoid them. Distinguishing between dynamic and static obstacles is a difficult task. The new technology of frequency modulated continuous wave LiDARs offers new options for this. In this paper, a system setup is presented with which dynamic objects can be detected, tracked and their dimensions, states and classes can be estimated on the basis of LiDAR sensors. Furthermore, the static objects are represented as convex polygons and merged with polygons from electronic nautical charts. The data and the information about the dynamic objects can be used by subsequent algorithms to plan a collision-free path. The presented method was successfully tested on an experimental catamaran equipped with 4 LiDAR sensors.