Fall detection for elderly using Android mobile phones

Abstract

Fall is the leading causes of injury among the elderly people. A falling accidence can cause the elder get serious health threat such as fracture, head traumas, and even death. In this project, a detection fall system is designed to give a quick response to seek help if a fall is detected by using Smartphone. By taking the advantage of prevalence and convenience of Smartphone use, the design of detection sensors based on threshold-based tri-axial accelerometer fall detection algorithm. In order to increase the fall detection accuracy, two body tests are taken to define thresholds for different users; accelerator and orientation sensors inside Smartphone also are used to this project. An alarm will go off when a fall action is recognized. Afterwards, a series of alert actions are triggered. I.e. an automatic call service, SMS, GPS location is sent to the emergency contact person set by the user. These detecting and alarm services were examined and integrated into an Android application with user friendly interface. The results suggest that there is more or less some false alarms feedback when detecting fall. Also there is a concern about enabling application running in background and power consumption. A further fall study is highly recommended to define individual fall thresholds.