We increased the accuracy of an instrumented treadmill’s measurement of center of pressure and force data by calibrating in situ and optimizing the transformation between the motion capture and treadmill force plate coordinate systems. We calibrated the device in situ by applying known vertical and shear loads at known locations across the tread surface and calculating a 6 × 6 calibration matrix for the 6 output forces and moments. To optimize the transformation, we first estimated the transformation based on a locating jig and then measured center-of-pressure error across the treadmill force plate using the CalTester tool. We input these data into an optimization scheme to find the transformation between the motion capture and treadmill force plate coordinate systems that minimized the error in the center-of-pressure measurements derived from force plate and motion capture sources. When the calibration and transformation optimizations were made, the average measured error in the center of pressure was reduced to approximately 1 mm when the treadmill was stationary and to less than 3 mm when moving. Using bilateral gait data, we show the importance of calibrating these devices in situ and performing transformation optimizations.
Goldberg is with the Department of Engineering, Hofstra University, Hempstead, NY. Kepple is with Department of Health, Nutrition, and Exercise Sciences, University of Delaware, Newark, DE. Stanhope is with the Department of Health, Nutrition, and Exercise Sciences and the Department of Mechanical Engineering, University of Delaware, Newark, DE.