Nursing home operators may soon be able to closely track resident movements using relatively few cameras, according to researchers at Carnegie Mellon University.
The team has taken multi-camera, multi-object tracking to the next level by creating a monitoring system that works outside a controlled lab environment. The team tested their system in a nursing home, where camera views were “compromised by long hallways, doorways, people mingling in the hallways, variations in lighting and too few cameras to provide comprehensive, overlapping views,” according to the university.
By using an advanced algorithm that takes advantage of facial recognition technology, the nursing home system had an 88% success rate for locating residents within one meter of their actual position. This is compared to 35% and 56% success rates for other algorithms.
“The goal is not to be Big Brother, but to alert the caregivers of subtle changes in activity levels or behaviors that indicate a change of health status,” said researcher Alexander Hauptmann, Ph.D., principal systems scientist in the CMU computer science department.
The researchers have compared their system to the fictional “Marauder’s Map” that gives Harry Potter and his friends the power to see where people are in the Hogwarts School of Witchcraft and Wizardry. But the team is also looking for ways to build privacy into the system, such as by identifying people by outline only.
The team will present their findings June 27 at the Computer Vision and Pattern Recognition Conference in Portland, OR.