2014 Fiscal Year Annual Research Report
最低センサー数を用いた人間の動さ計測と認識とその応用
| Project/Area Number |
14F04768
|
|---|
| Research Institution | Tokyo University of Agriculture and Technology |
|---|
Principal Investigator |
VENTURE Gentiane 東京農工大学, 工学(系)研究科(研究院), 准教授 (30538278)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
BONNET Vincent 東京農工大学, 工学(系)研究科(研究院), 外国人特別研究員
|
|---|
| Project Period (FY) |
2014-04-25 – 2017-03-31
|
| Keywords | 人間力学 |
|---|
| Outline of Annual Research Achievements |
Quantification and identification of human kinematic and kinetic variables using low-cost sensors: This module will receive as input the data from low-cost FPs, IMU, similar to the one embedded in recent smart-phones and/or of a Microsoft Kinect sensor. A RGB-Depth camera tracking the IMU position will be used punctually to cancel the drift. Also, merging an IMU and a RGB-Depth camera data will allow the motion capture system to handle occlusion that can easily appears during complex industrial tasks. New adaptive filters, based on Kalman filter and Weighted Fourier linear combiner, including a constrained kinematic model of the investigated limb(s) as state vector will be developed to fuse these two sensors.
The FP will be used prior to the measurement to perform a 3D fast identification of the inertial parameters of the subject.
|
| Current Status of Research Progress |
Current Status of Research Progress
1: Research has progressed more than it was originally planned.
Reason
The candidate is extremely brilliant and autonomous and he pursue research with a lot of enthusiasm. He also work in collaboration with students from the lab which allows to accelerate and produce excellent results.
|
| Strategy for Future Research Activity |
Quantification and identification of human kinematic and kinetic variables using low-cost sensors:
This module will receive as input the data from low-cost FPs, IMU, similar to the one embedded in recent smart-phones and/or of a Microsoft Kinect sensor. A RGB-Depth camera tracking the IMU position will be used punctually to cancel the drift. Also, merging an IMU and a RGB-Depth camera data will allow the motion capture system to handle occlusion that can easily appears during complex industrial tasks. New adaptive filters, based on Kalman filter and Weighted Fourier linear combiner, including a constrained kinematic model of the investigated limb(s) as state vector will be developed to fuse these two sensors. The FP will be used prior to the measurement to perform a 3D fast identification of the inertial parameters of the subject.
|
