Body Posture Detection Technique Based on MEMS Acceleration Sensor
Cheng Yang · Nanotechnology and Precision Engineering · 2010
In this paper,a body posture detection technique based on MEMS acceleration sensor was introduced.The signal vector magnitude(SVM) and the mean absolute value of differential SVM(MADS)of the body were used to describe the state of human motion and judge the state of fall.Multistage detection was applied to guarantee the real-time level and accuracy of the fall detection algorithm.First,suprathreshold SVM was detected as the elementary detection.Then,suprathreshold MADS was detected as the definitive fall criterion.After that,the acceleration variance within a certain period of time was calculated to estimate the stability of the posture to avoid misjudgment.Last,the angle between trunk and horizontal plane was calculated to ascertain posture,which would be sent in alarm message as the subsidiary information.In the system,the MEMS acceleration sensor was used to monitor body acceleration,fall detection algorithm was used to identify posture,GPSOne technology was used for orientation and SMS for alarm.Experimental results show that the algorithm is high in accuracy and satisfactory in real-time performance.