Feature - Working the Camera: Real-Time Grid-Powered Surveillance
Two people happen to meet in a hallway. They begin to chat, and end up in deep discussion.
Suddenly, on the wall beside them, projected images appear. Video clips. Sounds. Information directly relevant to their discussion. And all made automatically available, where and when they need it most.
This is the vision for the University of Kentucky's Ambient Virtual Assistant, a real-time grid application framework for managing grids that combine multiple sensors and output devices.
"You could think of AVA as a large-scale surveillance system, but it's much more than that," explains Hank Dietz of the University of Kentucky's Center for Visualization and Virtual Environments.
"The concept is to have the ambient environment constantly monitoring what's going on, not just recording, but intelligently reacting to what it sees and hears."
Sounds a bit space-age? It still is.
But Dietz says the AVA testbed is already underway, aiming to integrate many dozens of digital cameras, microphones, speakers, projectors and other sensors to allow real-time control and processing.
"Although prototypes of such applications have been built in the center before, our grid research centers on creating the real-time infrastructure that will allow these kinds of applications to scale up to using many different types of devices in larger areas."
One of the AVA's first test subsystems was "an old Nikon 950 digital still camera with a 185-degree fisheye lens," explains Dietz. This Nikon is the camera used to take the image above.
"Under computer control, that camera alone has captured approximately two million images. The grid infrastructure determines when to capture images and controls the capture parameters, such as exposure, resolution, frame rate, focus, lighting, timestamping, etc."
The AVA project is funded by the US National Science Foundation's Experimental Program to Stimulate Competitive Research.
- Cristy Burne, iSGTW