Motion Capture System
The Vicon Motion Capture (MoCap) system captures and records the movements of human models in the virtual reality CAVE. We have used this technology to replicate the movements of an orchestra conductor in real time, and to transfer the motions of a human dancer to a computer generated avatar.

Description
Motion capture technology is a vastly growing field with applications in everyting from human factor studies to feature films.

The subject wears a lycra suit equipped with small, ball-shaped reflectors. Each reflector consists of retroflective tape from 3M wrapped in a ball and strategically placed on the subject. 



A set of eight overhead cameras then tracks these reflectors as the subject moves within the covered cubic volume.

A computer communicates with each camera and cross references the data to calculate the position of each marker in space.

The cameras must be properly calibrated (right) and each marker must be visible to at least three cameras at all times.
The resulting 3d data is referred to as a point cloud, each point representing one reflector. This point cloud mirrors the 3d movements of the subject with great accuracy and is recorded for future use.

This data set can be used to animate virtual characters based on the motion of a real person.

For a more detailed introduction to Motion Capture, review this Powerpoint presentation (~4MB).
 
  
Specifications
Camera
Vicon Motion Capture MX13 Cameras
Supplier
Resolution
1.3 million pixel digital CMOS sensor
Sample Rate
484 fps, full frame
Configuration
8 cameras in radial configuration
Reflective Material
3M retroflective tape
 
  
Usage

Eric Maslowski and Steffen Heise supervise student and faculty usage of the Motion Capture system.

For questions, or to arrange an appointment to use the UM3D Lab's MoCap system, email: Eric or Steffen.

 
  
Gallery
Camera calibration
A Quicktime animation (~4MB)
Andy Kirshner
Michael Nickens
Copyright 2008 ©, the Regents of the University of Michigan, Ann Arbor, Michigan, USA