A wireless real-time fall detecting system based on barometer and accelerometer

Falls threaten the health and life safety of the elderly seriously. And timely medical assistance will help largely to reduce the rate of morbidity- mortality. In this research, a fall detection algorithm was proposed to detect various fall scenarios with high accuracy by analyzing the fall events data obtained from young subjects and the Activities of Daily Life (ADL) data collected from the elderly subjects. Meanwhile a wireless real-time fall detecting system embedding a tri-axial accelerometer, a barometer and ZigBee protocol was developed. The experiment results showed that the real-time system had portability and high reliability on fall detection with a sensitivity of 95.71% and a specificity of 97.78%, for a total set of 320 movements. The system is going to be put into practice on the old subjects in a geracomium in Beijing.