We developed an automatic method to extract the parameters of the 7 x 50 m Stroke Efficiency Test for swimming based on a wrist worn acceleration sensor device. In the wrist acceleration signal we detect characteristic swim events such as wall push-offs, wallstrikes and strokes. Based on this information we compute the distance per stroke and the swimming velocity. The upper error bounds of our automatic method are 1.67% for the velocity and 1.33% for the time per stroke. The velocity measurement accuracy is of comparable order to the manual accuracy. The automatic method clearly outperforms the manual measurement for the time per stroke extraction.
© Copyright 2008 Third International Conference on Pervasive Computing and Applications, 2008. ICPCA 2008. All rights reserved.