The real-time monitoring of human movement can provide valuable information regarding an individual’s degree of functional ability and general level of activity. This paper presents a fall-detection technology based on tri-axial accelerometer sensor was introduced. The acceleration data of the activities with the characteristic quantity SVM and SMA was analyzed. A method based on SVM and SMA, that takes the human body activity (erectness/lies down) as the auxiliary criterion to distinguish fall and ADL is proposed, and concrete threshold value and related parameters are summarized in this article. The experiment results proved that this scheme can obtain highly rate of accuracy and this algorithm has very good timeliness.