Inertial measurement unit (IMU) sensors are widely used in many wearable systems for detecting falls in real-time and reducing the risks associated with the fall in all instances of accidents. This letter presents a highly portable and comfortable IMU-sensor-based fall detection system for elderly people. Using an accelerometer and a gyroscope, the proposed system can monitor the sudden rise of the acceleration and the orientation of the user's head to detect accidental falls. When an accidental fall is detected, the proposed system transmits alarm messages to the data server via the wireless network. Also, a complementary filter is adopted to eliminate the instability and drift of angular measurement by the accelerometer and the gyroscope, respectively. The functionality of the proposed system is verified in experiments that involve 120 falling and 450 nonfalling actions of five participants. The experimental results indicate that the proposed system achieves a fall detection accuracy of 95.44% and has the potential to assist in the everyday healthcare of elderly people.
All Science Journal Classification (ASJC) codes
- Electrical and Electronic Engineering