다운로드 수: 43
업데이트 날짜: 2022/3/18
Supervised by Prof. Tung-Wu Lu, Dept. of BioMedical Engineering, NTU
MATLAB programs, which use the data derived from a motion capture system and force plates, for human motion analysis, including motion tracking, evaluation of body balance, and derivation of joint forces and torques of the lower body, etc.
- MATLAB 2015 (or newer versions)
- MTIMESX function(Fast Matrix Multiply with Multi-Dimensional)
The transformation between global & local coordinate of marker position on lower body segemnts.
Derive the COP from data of two force plates and display the positions relative to force plates.
- Represent the rotations of body segemnts with Euler angle and fixed angle.
- Display the difference of Euler angle before and after static calibration of indivisual.
- Smoothen the data curve and display the angular velocity of lower body.
- Consider right foot only, compare the results of analytic solution of angular acceleration & 1st derivation of angular velocity.
- Consider right thigh only, compare the angular velocity derive from 12 sequence of Euler angle.
The one with Gimbal lock during motion can be easily observed and then be avoided.
- Write the function "unwrapEP.m" to eliminate the discontinuity of Quaternions data.
- Compare angular velocity & angular acceleration derived from Euler angle and Quaternions(EP).
- Compare the rotation axis and angle derived from Screw axis and Quaternoins.
- Determine the joint center & rotation axis with least-square error from several rotation axes dervied during the motion.
- Derive & compare the COM position of whole body using Dempster's anthropometrical data with simplifying the body model as 7, 11, 12, & 13 segments.
- Display the COM positions. (Yellow: markers' position, green: COM of body segments, purple: COM of whole body).
Derive the COP data during the recipient is standing still. Determined the eclipse covering 95% of these COPs with Principal COmponent Analysis (PCA). The area of the eclipse and length of its axes can be a indicator for evaluating one's body balancing ability.
Derive the joint moment & joint force of angular momentum during the motion. The results are normalized by being divided by the weight of recipient.
In the last few minutes of Marvel's The Punisher S1E1 (Netflix), a similar result appeared when David Lieberman used gait detection and recognized Frank Castle, who faked dead after revenging for his family.
Wei-Hsiang Wang (2022). Human-Motion-Analysis (https://github.com/mattwang44/Human-Motion-Analysis-MATLAB/releases/tag/0.0.1), GitHub. 검색됨 .