Video based movement measurement systems are commonly used during functional movement analysis to determine the three-dimensional (3D) locations of targets placed on the subjects being measured. The 3D target locations are tracked over time to generate target trajectory time histories, which are used to determine joint positions over time (joint angle time histories). Unfortunately, some of the technological advances to make the movement measurement systems more automated can jeopardize the accuracy of target trajectory data, and thus the interpretation of that data. Target trajectory artifacts may not always be adequately dealt with through standard noise removal techniques (eg. low pass filtering, polynomial curve fitting, least squares techniques). The purpose of this project is to develop, evaluate, and demonstrate a new target trajectory quality indicator. Three trials of target trajectory data were collected using a sampling frequency of 50 Hz as a subject walked while swinging an idealized lower extremity model next to her right leg. The leg model to which the targets were attached was rigid, so that any change in intertarget distances was an indication that a measurement error had occurred. Computer software were used to generate target trajectory time histories, joint angle time histories, and various quality indicators. The absolute value of the sample to sample change in acceleration (the delta A parameter, or DAP) was chosen as the new target trajectory quality indicator because it was the only one, among the differentiation techniques that were considered, that uniquely identified the location and relative magnitude of each of the idealized target trajectory artifacts. The DAP was computed by calculating the change in acceleration between one target trajectory data point and the next. The translational acceleration of a target trajectory data point was determined using a centered finite divided difference approach.
