CALL FOR PAPERS


The vast majority of the Earth is covered by water, with the surface, water column, and bottom of these water masses all providing habitat for organisms in addition to hosting dynamic physical, chemical, and geological processes. However, most of the water environment is understudied, despite the critical importance of the ocean for Earth’s biodiversity and climate. Automated analysis of underwater imagery has become a vital tool in the study of marine and freshwater environments, particularly in the deep sea, where other study methods are impractical. Computer vision approaches enable the analysis and understanding of huge amounts of data, through which new insights can be gained which are unreachable with traditional imaging and manual annotation methods. Hence, automated analysis of underwater imagery, for both fixed and mobile camera systems, is an emerging and rapidly growing field.


Underwater imagery can be collected through a variety of means. Imagery is often acquired by mobile camera systems which are deployed by ships via towed systems, Remotely Operated Vehicles (ROVs) or their autonomous counterparts (AUVs). These methods offer only a limited temporal sampling of complex underwater phenomena, but can provide broad spatial coverage reaching a diversity of environments. Oceanographic long-term, high-resolution data acquisition has been greatly facilitated by the establishment of seafloor cabled observatories whose co-located sensors allow for interdisciplinary studies and real-time observations augmenting traditional oceanographic research approaches. These seafloor cabled observatories, such as those operated by Ocean Networks Canada (http://www.oceannetworks.ca), offer a 24/7 presence, resulting in unprecedented volumes of visual data.


All underwater imagery imposes a series of unique challenges for computer vision algorithms including but not limited to, sparse access to the study environment, harsh conditions (high pressure/low temperature/currents), imaging within a scattering medium, limited and/or artificial lighting, and only indirect or no access to global positioning data. These challenges must be tackled by the computer vision community in collaboration with physicists, biologists and ocean scientists. To this end, we invite submissions from all areas of computer vision and image analysis relevant for, or applied to, underwater imagery. Topics of interest include, but are not limited to: