The fall detection method based on the wireless acceleration sensor

To judge whether the alone elderly fall or not is an important need in the elderly health supervision. This paper puts forward a method based on three axis acceleration data and pattern recognition has been presented to judge the situation of falling for old men. The method was based on an acceleration transducer named MMA7260Q. Due to the characteristic that the three-axis signal divers in a huge area during the progress of old men falling, it combined the peak value of acceleration and acceleration energy curve to proceed three axis acceleration data for detecting the falling, it could avoid outputting errors due to the changing of angle of old men. According to the result of the experiment, the accuracy rate of the method of detection proposed in the report could run up to 93 percent.