Hybrid vision-based motion capture for a human leg

Studenteropgave: Speciale (inkl. HD afgangsprojekt)

  • Miguel Gargallo Vazquez
4. semester, Vision, Grafik og Interaktive Systemer, Kandidat (Kandidatuddannelse)
The scope of this project is: computer vision and computer graphics, where the focus is on motion capture.
Marker-based and marker-less motion capture systems present a opposing set of advantages and disadvantages in terms of convenience and accuracy that could be made use of by combining them in a hybrid system.
The aim of the project is to provide the necessary background knowledge and a methodology to find a combination of both systems that would result in a more convenient set up.\\
An independent implementation and tests are carried out in a simplified scenario where a human leg moves and walks along a line. It is recorded with a camera and a marker-based system, using a set of markers attached to the different segments of the leg. The segments obtained in both systems are combined in different ways and the angles for the knee and ankle are computed and compared to those from the full marker set.
SprogEngelsk
Udgivelsesdato8 sep. 2017
Antal sider30
ID: 262328106