Computer vision for reverse engineering in the design, simulation and operation of maritime systems

We present, here, a compilation of attempts to use computer vision technology in the maritime field. The scope is mostly in reverse engineering, that is, get- ting information about an existing physical system and modelling it digitally, extracting information from images and videos. The paper starts with a broad defi- nition of computer vision, and how it can be adapted to reverse engineering. It is followed by a modest lit- erature review of cases in the maritime sector, namely photogrammetry, inspection and. Later, it describes a recent project supervised by the first author, using commercial software to re-model digitally the hull of a small sailing boat. The use of open-source tools is later mentioned, exemplified by the research con- ducted by the second author on the use of cameras as sensors for acquiring dynamic data, closing the dig- ital twin loop. A proposal for the use of image data extraction is presented at the end, as a substitute for the expensive re-sensoring of vessels in operating, in the current hype of remote operation and digital twins. We conclude by emphasizing the potential of this tech- nology for the transition from conventional to remote and autonomous operations.