PT Journal
AU Zhang, L
   Brunnett, G
   Rusdorf, S
TI Real-time Human Motion Capture with Simple Marker Sets and Monocular Video
SO JVRB - Journal of Virtual Reality and Broadcasting
PY 2011
VL 8(2011)
IS 1
DI 10.20385/1860-2037/8.2011.1
DE Computervision-based motion capture; Human motion capture; Inverse kinematics
AB In this paper we present a hybrid method to track humanmotions in real-time. With simplified markersets and monocular video input, the strength of bothmarker-based and marker-free motion capturing areutilized: A cumbersome marker calibration is avoidedwhile the robustness of the marker-free tracking is enhancedby referencing the tracked marker positions.An improved inverse kinematics solver is employedfor real-time pose estimation. A computer-visionbasedapproach is applied to refine the pose estimationand reduce the ambiguity of the inverse kinematics solutions.We use this hybrid method to capture typicaltable tennis upper body movements in a real-time virtualreality application.
ER