Sleep-pose Recognition Based On Pyroelectric Infrared Sensing Technology

In this paper, a innovative method based on Pyroelectric Infrared sensing is proposed for sleep-monitoring. First, a pyroelectric infrared sensor is used to receive the infrared radiation emitted from a human body. Then we extract the motion information by analyzing the features of analog signal, which is outputted from the pyroelectric infrared sensor. At last, the nearest neighbor algorithm based on modified Hausdorff-distance is employed to achieve the sleep position recognition. The experimental results show that the biggest correct classification rate of this method reaches 83.55%. This new method provides a nonintrusive way of sleep position recognition for sleep-monitoring.