Latest question:
How do you think the new GigE standards will influence the machine vision industry?
Respond or ask your question now!


By Leonard A. Hindus
Contributing Editor
MBARI Software Engineer Brian Schining says the VARS (Video Annotation and Reference System) consists of three parts, the annotation system, the knowledge base and the query system.
The Annotation System
The annotation system is how data gets entered into the knowledge base. The annotators scan the tapes looking for biological or geological features. Each animal and feature is classified. When a feature or animal is entered, the system notes the time code from the tape or disc. At this stage, these data are matched with conductivity, salinity, temperature, depth, oxygen and navigation data and ancillary data such as vehicle position, and camera state values including focus, zoom, iris and fieldwidth. The annotator has the option of grabbing and storing the frame and can enter descriptive information such as "The rockfish swam on the sea floor," or "The rockfish is red."
The user interface for the annotation system can be customized by each annotator. The user can set up a custom icon tab, allowing them to add observations. Each icon represents a word, specifically the name of a concept. For efficient annotation, the user can create separate tabs for different types of dives, such as Geology or Midwater, and populate each with the names of observed objects in those categories.
The Institute operates a number of remotely operated vehicles including the Tiburon and the Ventana. The Tiburon has a Panasonic WVE550 three- chip camera. The Ventana has a Sony HDC-750A high-definition camera. For recording all dives, they use Sony's DVW-500/A500 Digital Betacams. For recording select dives, a Panasonic AJ-HD2000 high-definition recorder is used. The annotators scan the images using a studio monitor and a Sony DVW5000 digital betacam tape deck.
"We save all our video to tape," says Nancy Jacobsen Stout, MBARI's video lab manager. "It is important to retain the maximum definition possible and only tape has sufficient capacity without any compression." Tape also lets them archive the video over time.