A map that tracks people’s movements across a network of CCTV cameras and gives their location in a real time could be an invaluable tool for finding intruders, or for simply knowing your whereabouts.
Dubbed the Marauder’s Map after the magical map used by Harry Potter, the system takes security camera footage and analyses it using an algorithm that combines facial recognition, colour matching of clothing, and a person’s expected position based on their last known location.
The main challenge in designing the map, says Shoou-I Yu, a graduate student at Carnegie Mellon University in Pittsburgh, Pennsylvania, was finding and following individuals in complex indoor environments where walls and furniture can block the cameras. He and his colleagues found a solution by combining several tracking techniques. For example, an individual whose clothes are the same colour, and whose facial features correspond to those of the person who appears just a few frames back in the footage suggests that the two are a match.
The team tested the map in two settings: an open outdoor scene and the public areas of a nursing home. In the nursing home, they monitored 13 people as they moved through the building over a 6-minute period. The map tracked people to within 1 metre of their actual position, and its algorithms were far more accurate than other software systems tested. It registered just a fifth of the false positives and false negatives of the next best system.
Yu claims that the Marauder’s Map is the first to track multiple individuals over many cameras in a complex indoor environment. Such a map could one day be used to find, say, a nursing home resident with dementia who has got lost, or to help guide people through large stores or shopping malls.
The researchers will present a paper on the work at the Computer Vision and Pattern Recognition conference in Portland, Oregon, in June.
Richard Radke of the Rensselaer Polytechnic Institute in New York says the Marauder’s Map solves on a small scale a challenge that he’s attempting to tackle in a much larger setting. He is working on building a system that tracks people flagged for monitoring by security officials at Cleveland Hopkins Airport in Ohio.
“There’s a lot of interest in this kind of human re-identification,” Radke says. “We have big networks of cameras – some overlapping, some not – and the idea is to locate someone on one camera and track them as they move around. The difference is that the airport has hundreds or thousands of people moving.”
(Article by Hal Hodson)