LiDAR and RGB Camera Performance for Obstacle Detection in Marine Environment

Environmental perception is a critical skill for autonomous vehicle applications in maritime environments. Two of the most important sensors employed for this task are RGB cameras and LiDARs. In this work, we present a data-driven performance analysis of the two sensors in maritime obstacle detection tasks. The aim is to assess the performance of each sensor and then evaluate the compatibility and redundancy between the two, across a comprehensive set of different maritime scenarios. This type of analysis is yet to be seen in maritime literature, partly because of the scarcity of public data. As additional contributions, we present publicly available annotations for some public datasets found in the literature.