Tri-axis accelerometers are widely used to detect physical activity. Three-axis accelerometers are sensing devices that measure gravitational changes. In this chapter, a body motion index was derived using an axis accelerometer with additional signal filtering and feature extraction. Various body motion factors, motion rates, motion angles, and directions were examined during an experiment. Six participants were recruited for this study. A TD1A system (K&Y Labs) was employed; this wireless system comprised one electrocardiogram (ECG) and tri-axis acceleration sensor. Using a belt, an amplifier was fixed to each participant in the same relative position between the abdomen and chest for each experimental measurement. The participants were instructed to move in both directions seven times for varying durations and at differing motion angles. Features were extracted from the motion index series. Both the mean and maximum values of the motion index series were used. The results showed that specific posture change patterns had corresponding axis acceleration variations. The influence of the motion angle on the motion index value was considerably greater than that of the motion rate. Higher motion angles were achieved with higher motion index values. Generally, anterior-posterior swaying caused greater motions than left-to-right swaying under the same motion conditions for angle and rate. Therefore, the proposed motion classification algorithm combined with a tri-axis accelerometer has significant potential for motion detection. However, the use of accelerometers has a number of limitations. In the future, a multi-sensor system will be employed to detect body movements.
Relation:
Lecture Note in Electrical Engineering,234,147-151.
