Uni boffins create life-saving search and rescue system which differentiates how people move
A system which reduces the amount of time taken to locate missing people on open ground has been developed by scientists from the University of Wroclaw.
The Search and Rescue Unmanned Aerial Vehicle (SARUAV) produces images taken from unmanned aircraft to automatically assist in search operations by showing rescuers potential sites on map reports.
Created by a team led by Prof. Tomasz Niedzielski from the SARUAV technology company working alongside the University of Wroclaw, the system first identifies an area where the missing person might be.
Prof. Niedzielski said: "It does that with the help of several mathematical models, which take into account the specifics of the person who is missing - because a child, an ill person, a tourist and a cyclist all move differently - as well as data about the coverage and shape of the terrain.
"A person moving by road usually moves more quickly than someone tackling heavy undergrowth, woodland or marshland."
Drones are then sent on a mission only over those identified areas of land to take aerial photographs, on which a special algorithm identifies people by looking for specific anomalies in the picture, including in colour, which are then checked for similarity with defined characteristics of the person.
At the same time, the system can mark the coordinates of identified places, which enables the rescue services to get to the site of the missing person quickly.
The phase of narrowing down potential sites takes a dozen or so minutes and the team's priority is to reduce the time it takes to process the pictures taken by the drone.
"Our first trial ended in 20 minutes on one picture, now we've got it down to single-digit minutes," Niedzielski explained. "We're doing two things at once: were cutting the time of calculations, and we're also trying to increase the effectiveness of detection."
The system should be ready for use in search operations within two years.