Computer Vision
Harvesting the thermal cardiac pulse signal
MICCAI'06 Proceedings of the 9th international conference on Medical Image Computing and Computer-Assisted Intervention - Volume Part II
Human-robot interaction in rescue robotics
IEEE Transactions on Systems, Man, and Cybernetics, Part C: Applications and Reviews
Hi-index | 0.00 |
This paper presents two new algorithms for measuring human breathing rate automatically: a Binary Algorithm and a Histogram Cost Algorithm. These algorithms analyze frames from a thermal video of a person breathing and then estimate the person's breathing rate. Our Binary Algorithm reduces grayscale images into pure black and white (binary) images. Our Histogram Cost Algorithm enhances the differences on normalized histograms by assigning a larger cost to darker pixels. We tested our algorithms on 26 human subjects and results show that the Binary Algorithm's total percentage error is 19.50%, while the Histogram Cost Algorithm's total percentage error is 4.88%. These algorithms work in real time, presenting constantly updated measurements of the breathing rate. They are also resistant to small movements and work under several environment conditions, which makes them suitable for measuring victims' breathing rate in Urban Search and Rescue Situations, as well as patients in Medical Situations.