Vol.7 No.4AA115-NT74-75 |
|||||
Academic Articles | |||||
Regular Paper | Vol.7 No.4 (2016) p.206 - p.215 | ||||
POD Evaluation for Joint Angles from Inertial and Optical Motion Capturing System |
|||||
Kai SHIMIZU1, Futoshi KOBAYASHI1,*, Hiroyuki NAKAMOTO1, and Fumio KOJIMA1 | |||||
1Dept. of Systems Science, Kobe University, 1-1 Rokkodai-cho, Nada-ku, Kobe 657-8501, Japan |
|||||
Abstract | |||||
It has been recognized that advances in preventive maintenance can improve the sustainment of systems, facilities, and infrastructure. Robot technologies have also received attention for maintenance applications. In order to operate delicate tasks, multi-fingered robot hands have been proposed in cases where human capability is deficient. This paper deals with motion capturing systems for controlling the hand/arm robot remotely. Several types of motion capturing systems have been developed so far. However, it is difficult for individual motion capturing systems to measure precise joint angles of a human arm. Therefore, in this paper, we integrate the inertial motion capturing system with the optical motion capturing system to capture a human arm posture. By evaluating the reliability of each motion capturing system, the integration is carried out. The probability of detection (POD) is applied to evaluate and compare the reliability of datasets measured by each motion capturing system. POD is one of the widely used statistical techniques to determine reliability. We apply the analysis to determine the POD() function from the data set. Based on the POD evaluation, two motion capturing systems are integrated. | |||||
Keywords | |||||
Robot Maintenance Motion capturing system, Optical sensor, Inertial sensor, Tele-operation, Probability of detection | |||||
Full Paper: PDF
|
|||||