ABSTRACT
Indexed by Track Subject
Indexed by Title
Indexed by Author
 Geopositioning a Remotely Operated Vehicle for Marine Species and Habitat Analysis
Track:  Oceanography, Coastal Zone, Marine Resources
Author:   Paul Veisze
Konstantin Karpov
The immense species richness and cryptic nature of many marine benthic organisms require years of training to qualify scientific investigators. A more viable option for producing quality data may be to train (relatively abundant) divers on how to obtain quality video images for use by (relatively scarce) personnel trained in algal and invertebrate identification. With this general goal in mind, the California Department of Fish and Game Marine Region has mobilized research at the Punta Gorda Ecological Reserve (PGER) (Humboldt County, California), which includes development of noninvasive observation techniques using a remotely operated vehicle (ROV). A Deep Ocean Engineering Phantom HD2 ROV, equipped with a video camera and two parallel laser beams spaced 10 centimeters apart, was deployed for measurements of macro invertebrates and finfish. Concurrently at timed intervals, the ROV was landed for random recording of fixed-area quadrates, recorded both perpendicular to the substrate with the rotating camera facing downward and horizontally in the case of vertical reef faces. Geopositioning of the ROV was estimated by deploying a Trimble GeoExplorer II GPS receiver in a waterproof float tethered to the umbilicus of the ROV. A total of 1,100 position fixes were recorded along an ROV transect at a depth of 30 meters during a one-hour deployment. ArcView GIS tables of UTC time and position from the float-based GPS (postdifferentially corrected) were linked to similar tables generated from a shipboard recording of the ROV video camera, which also had GPS-based UTC time code. This technique enabled estimation of the ROV position for any of the video frames of marine organisms captured from the recording. The ArcView GIS hot link function was then used to display the selected video frames upon spatial selection of ROV positions. DFG has immediate plans to deploy the ArcView Tracking Analyst extension for real-time visualization of ROV positions. This will involve acoustic-based tracking of the ROV's position relative to the surface vessel's real-time DGPS position. Seafloor image classification will then be ground-truthed using the ROV video camera, with the ROV position displayed over the subject seafloor images.

Paul Veisze
State of California
Information Technology Branch GIS Unit
1807 13th Street
Sacramento, CA95814

Telephone: 916-323-1667
Fax: 916-323-1431