Research Of Human Movement Monitoring Based On Accelerometer

With the further development of microelectronics and sensors，the sensors can be embedded in awide field of mobile phones, portable devices, etc. It has broad application prospects that useacceleration sensors to monitor human movement. Human daily movement and personal health areclosely related. On the one hand, monitoring of human daily movement is significant for people’splan of scientific exercise to benefit one’s physique. On the other hand, monitoring of humanabnormal movement has very practical value. Fall brings a great threat to human body. Thereal-time detection can provide timely help and reduce risk.Monitoring of human movement generally brings large amount of calculation and long responsetime with multiple sensors. This thesis proposed a mechanism of human movement recognitionbased on a single acceleration. The main work includes:1. By analyzing and Summarizing the theory and research methods commonly used in the fieldof the monitoring of human motion which is based on acceleration sensor, the paper analyzevarious modules such as data acquisition, preprocessing, feature extraction, selection andclassification in process of statistical pattern recognition.2. This thesis proposed a human action recognition mechanism based on a single accelerationsensor. The mechanism calculates the vertical direction eigenvalues of the threshold before the crest,the threshold before the trough and the interval between peaks and valleys, and combines withdecision tree classification algorithm, in order to identify the action of squatting, jumping andrunning.3. This thesis proposed a human fall detection algorithm based on a triaxial accelerometer. Thealgorithm effectively distinguishes four kinds of behavior falls which are front falls, rear falls, leftfalls and right falls according to extracting human motion feature and calculating three axes. Theexperimental results show that the mechanism of human action recognition correct rate reachesmore than90%, proving the practical application value of this study.4. This thesis designed a human movement monitoring system. The functional structure of thesystem was introduced in detail.