Triaxial Accelerometer Fall Detection Technology-based Research And Application

With aging population becoming a worldwide problem, how to realize effective and reliable elderly people-oriented remote medical treatment and activity detection system has become one the hot topics in the research field of Wireless Sensor Networks.The main goal of fall detection system is to distinguish falls and some other activities in daily life, detect falls accurately and give alarms if necessary to minimize fall injury.Traditional solutions include video image analysis method and audio signal analysis method which both has some limitations. However wearable detection device is smaller and easy to carry with, it also has low power consumption, which meet the requirements of fall detection system. The device collects striaxial accelerometer data through microcontroller unit and transmits them wirelessly for real-time monitoring in order to give accurate fall alarms.To address these problems, this paper has made some valuable researches of fall detection system based on MEMS(Micro-electro-mechanical System). In this paper, we use tri-axial accelerometer MMA7455and platform CC2430to constitute the data acquisition node, triaxial accelerometer data are transmitted to sensor gate node through radio frequency then to PC through serial port, the original data of unintentional fall and daily behavior are collected. This paper also reduces the noise generated error through the methods of hardware compensation and Kalman filtering. Meanwhile a new three-level fall detection algorithm mainly based on SVM (Signal Magnitude Vector) is proposed. Finally we use C#language and visual studio2010development environment to realize fall detection system on PC, The simulation results demonstrates the feasibility and effectiveness of the proposed method.