A SVM algorithm for falling detection in an IoTs-based system

Abstract

Falling of elderly people is one of main reasons causing serious injuries or the risk of early death. However, this may be reduced by using an IoTs-based fall detection system, in which a SVM algorithm and PCA features are applied. In addition, datasets collected from tri-axial accelerometer sensors and/or Kinect camera systems are transferred to a central Hub via Zigbee interface and are updated continuously to a cloud server for processing and detecting fall states. In addition, fall messages can be sent to relatives through smartphones and/or healthcare centers for alert and supporting soon. Experimental results show to illustrate the effectiveness of the proposed system.