Photogrammetry for digital hand surface capture in swimming performance study

Abstract

The paper describes an application of close-range photogrammetry in the study of the human hand for propulsion in competitive swimming. The study involved the capture of 3-D surface data and modelling of a real and a replica hand. Both real and replica models were used in computer simulation to determine the propulsion parameters for high performance swimming. Custom-made control frame was used to provide accurate object-space control of the stereo-photography. In addition, the control frame was used to connect both sides of the hand into one coordinate system. A two-camera and a four-camera configuration were developed for the photography. The latter was used in the real hand photography and under-water stereo-photography. Results of the study showed that the surface of the replica has an undulating roughness of about 1.0 mm. However, the total error of the computed surface area was less than 1.5 cm2 or 0.85%. Consequently, the error of the surface topography of the replica was negligible and the resulted error in the computation of the propulsion parameters was minimal