Internship: Integrating multiple cameras in an Unmanned Survey Vessel

At DEMCON Unmanned Systems several boats have been developed that are able the sail a planned mission based on RTK-GPS data. These Unmanned Survey Vessels (USVs) will be extended with collision avoidance strategies based on LiDAR data in the near future. The vessels will then be able to autonomously avoid obstacles using data from the LiDAR. It is also desirable to extend the sensor suite by multiple cameras so that a 360 view can be obtained around the boat. This camera data can then be used for avoiding obstacles as well.

During this internship the student will design a system to incorporate cameras into the sensor suite. The images from the camera should be acquired and stitched together. The stitched image is then to be converted to a Robotic Operating System (ROS) message so that is can be published to other system nodes. The next step is to calibrate the data of the stitched image to the LiDAR data so that each LiDAR point can be assigned a colour value, which is convenient for filtering, segmentation and classification purposes. It is important to design the camera system and accompanying software in such a way that the cameras can be replaced in a modular fashion.

The student will have to design, implement and test autonomous functionalities on embedded systems. During the implementation phase the student will have to program in C++ and experience with this programming language is therefore desirable.