Markerless motion capture system via kinematic analysis of angular lower limb

Khairu Razak, Siti Badriah (2022) Markerless motion capture system via kinematic analysis of angular lower limb. Masters thesis, Universiti Tun Hussein Onn Malaysia.


Download (602kB) | Preview
[img] Text (Copyright Declaration)
Restricted to Repository staff only

Download (3MB) | Request a copy
[img] Text (Full Text)
Restricted to Registered users only

Download (29MB) | Request a copy


The introduction of markerless sensor technology in motion capture system offers a comparable alternative to the conventional systems by employing infrared-depth sensors and retaining the ability to acquire two (2D) and three-dimensional (3D) data on human movement. However, its accuracy is often questioned compared to the established technologies such as passive marker systems. Therefore, this study sets an alternative method to evaluate Kinect Xbox 360 markerless system accuracy based on two positioning coordinates of two pairs of sensors. Through this approach, the length of lower limb segments was measured in 2D and 3D on each motion frame while performing squat movement and compared with the actual segment length. Interestingly, all segment lengths in the 3D showed excellent accuracy with the actual length of the segment. The angle of knee joints was also evaluated to identify the types of squat movements. The same evaluation is also used for the accuracy of a passive marker system while capturing the turning kick motion. In addition, the velocity of the knee joint was also studied at each phase of movement to determine the speed and angular of the knee required to enable the subject's foot to reach the target. For validation purposes, simulations of all recorded motions were implemented to evaluate the squat and the phases of movement in a turning kick from a visual angle. Successfully, the study was able to compare the accuracy and precision of the system constructed using lower limb data relative to the passive marker system using actual lower limb data. The markerless gave a remarkable difference value between the highest and lowest percentage coefficients of variation with 3.90%, while the passive marker system gave 5.72%. It is suggested that the multi-camera markerless motion capture system used in this study be used only for applications that do not require a significant level of accuracy such as animations, gaming and recreational sports analyses.

Item Type: Thesis (Masters)
Subjects: T Technology > TA Engineering (General). Civil engineering (General)
Divisions: Faculty of Mechanical and Manufacturing Engineering > Department of Mechanical Engineering
Depositing User: Mrs. Sabarina Che Mat
Date Deposited: 15 Feb 2023 07:21
Last Modified: 15 Feb 2023 07:21

Actions (login required)

View Item View Item