Seafloor photo-imaging: challenges, methods and applications

The deep sea is a challenging environment to explore. In particular, low visibility in the deep-sea constrains images of the seafloor to being taken from a very short distance (<10 m) to the bottom. Such low altitude severely limits the field of view of individual images. Hence, mapping an entire site of interest requires many images. Generating robust photo-mosaics of the seabed is a key step towards getting a better knowledge of the deep ocean communities and geology. However, geo-referencing still images taken from remotely operated vehicles (ROV) for integration into a globally registered map constitutes a main challenge as the underwater environment poses numerous technical and environmental uncertainties. Such difficulties can be partially overcome by combining image analysis techniques, such as automatic feature detection techniques, together with available positioning data in order to constrain registration errors. In this presentation, I will also introduce such a tool for large-area photo-mosaicking and present application examples at cold seep and hydrothermal vent systems.