Detection of Human Body Movement Patterns Using IMU and Barometer

With the 20th century's drastic increase in the human population, population aging is becoming a major cause for concern. With current organizations and service staffs struggling to adequately care for an aging population, there is an increasing need for the use of technologies that can assist with this care. Human motion detection technology, which detects and reports on movements of the human body, can offer solutions to these issues. In general, motion detection technologies are divided into two categories: visual detection and sensor network detection. This paper proposes the combined use of an inertial measurement unit (IMU) and a barometer to conduct sensor network detection. The IMU device includes a three-axis accelerometer and a three-axis gyroscope used for body coordinates' accelerations and angular velocities. The barometer is used to measure pressure and temperature at various points. In this paper, the k-nearest neighbor (KNN) and the support vector machine (SVM) act as a complementary filter algorithm to detect four different human body movement patterns: the standing-up pattern, the falling-down pattern, the running pattern, and the walking pattern. The proposed algorithm was tested in a hardware and software platform, and the results have shown a classification accuracy of over 90% for these movement patterns.