Abstract:
Kinematic measurement is an integral part of biomechanics and the current gold standard is to use optical motion capture to track segmental positions and orientations. However, optical motion capture is limited to small capture volumes, and requires significant time and cost. One method to overcome these limitations is to use inertial measurement units (IMUs) to measure the linear accelerations and angular velocities of the body segments. Typical IMU-based motion measurement systems integrate acceleration and angular velocity data, resulting in numerical drift. Although data fusion algorithms have improved in recent years, reducing the error from drift, these methods are not immune to drift problems over longer time periods (i.e. > 20mins). This paper presents a new approach that combines IMU data with a capacitive stretch sensor to perform drift correction. We tested our approach using a simplified model of the elbow joint and illustrate how the addition of a stretch sensor can improve the estimation of a simple flexion-extension joint motion over an 80 min epoch.
