Shoulder-worn camera acts as a third eye - tech - 26 March 2007 - New Scientist Tech

Shoulder-worn camera acts as a third eye -

location based services

A shoulder-mounted camera system that automatically tracks head movements and can recognise hand gestures has been developed by UK researchers. Eventually, they hope the system could identify a wearer's activity and offer assistance, for example by accessing a telephone directory when they reach for the phone.
The collar-mounted camera, developed by Walterio Mayol Cuevas from the University of Bristol, is worn on one shoulder. It is controlled wirelessly from a laptop computer, which uses the camera's output to keep track of objects, map its position and recognise different hand gestures made by the user.
Three separate motors make the camera highly directional, and even allow it to tilt, while inertia sensors are used to keep it pointed correctly while the wearer moves around.
A video shows the the camera following the wearer's gaze using a second camera positioned elsewhere in the room. Another clip shows the camera rapidly mapping the location of different objects in a scene.
To the point
The system can to recognise when the wearer is pointing with a finger, grasping an object or resting their hands on a surface.
Pointing can direct the camera towards a target. Or it can be used to overlay virtual objects on a video picture. This video shows virtual objects being generated and positioned with different hand gestures.
"If you are going to use wearable computers, you cannot use a computer and mouse," Mayol Cuevas says. Instead, users may one day assemble their own "virtual workspace" around them using a wearable computer and a head-mounted display, he suggests.
The ultimate goal is to have the system understand what a person is doing simply by observing their hands. This should be easier than trying to decode everything happening within a scene, says Mayol Cuevas.
"A person's hands are always there to see," he told New Scientist. "I want to have the system use that to tell if you are, say, cooking or eating." This information could warn a computer to reroute phone calls to avoid interruptions, or cause it to bring up information on a computer screen or head-mounted display related to the task in hand.
Capture and go
"A wearable camera can remove the need to explicitly report what you are doing," says Chris Baber is researching wearable cameras at Birmingham University, also in the UK, who was not involved with the work.
Baber is working on another wearable camera system that could offer assistance to people such as police and archaeologists who spend a lot of time producing detailed reports about a location. Police in north England are currently testing a much simpler helmet-mounted camera system that can be used to record evidence.
"A camera can let you capture as much information as possible and assemble it into a report as you go along," he told New Scientist. This could cut the amount time spent filling in forms and make it easier for others to understand the situation.
Baber's system relies on voice commands rather than gestures. "Gesture recognition could be useful though," he says, for example recognising a specific task, like dusting for fingerprints.