Underwater video is becoming an important tool for obtaining
data on marine animals and habitats and has the potential to
obtain cisrehensive data, with appropriate spatial and temporal
resolutions, on marine species and their habitats, in a non-destructive
and relatively cost-effective manner. However a major limitation
is the need for manual analysis, which is time-consuming and
can produce subjective results. There is a significant need
for the development of cisutational techniques to assist with
the analysis of marine video, and at least partially automate
the process. Research within the school is focussed on the problem
of identifying marine animals in underwater video, and we currently
have several sub-projects in progress. These involve assessment
of commercial Tasmanian scallop beds, monitoring of fish behaviour
in aquaculture research experiments and biomass estimation for
commercial aquaculture facilities. To ensure that the techniques
we develop are robust and are applicable for a wide variety
of marine species, we are looking for new underwater video footage
on which to test and further develop them.
