A Comparative Analysis of Two Approaches for Estimation of Upper Limb Orientation Using Inertial and Kinect Sensors

Abstract

In Neuropsychology and Occupational medicine, the clinical environment uses upper limb assessment activity, which involves performing a few daily living activities. The main objective of this work is to develop a non-wearable, automated version of upper limb assessment, which allows free movement of the hand and records the trajectory information using inertial and Kinect sensor modules. The upper limb orientation is measured using an object attached to an MPU6050 inertial sensor. The assessment involves the measurement of quantitative parameters such as time, orientation, and trajectory with which the given task is completed. A similar task is performed using the Kinect sensor. The correlation between these two sensors is recorded. The Wilcoxon signed-rank test is performed to quantify the comparison between the stopwatch with the inertial sensor and Kinect sensor-based timing measurements.

Keywords

• inertial sensor
• kalman filter
• upper limb
• occupational medicine
• sensor fusion