|
Vision Based 3D Tracking with Applications to Augmented Reality
In augmented reality applications, tracking and
registration of cameras and objects are required because, to combine
real and rendered scenes, we must project synthetic models at the
right location in real images. We present here an accurate solution
that works in real-time even under difficult conditions. We also
worked on object detection.
3-D Face Tracking Demo
Our 3D tracking algorithm is robust enough to track a human face,
given a rough and generic 3D model of the face. Then virtual objects
can be added in a realistic way and in real-time. We use a 2-D face
detector to automatically initialize the tracking.
This demonstration was presented at different exhibitions where we
successfully tracked the faces of hundreds of people of all ages (some images).
Rigid Object 3-D Tracking
Combining Edge and Texture Information
We also integrated the ability to take edge-information into
account. We propose a method for handling cluttered or
textured backgrounds that is similar in speed to approaches that
consider only single hypotheses.
Relevant Publications
L. Vacchetti, V. Lepetit and P. Fua, Stable
Real-Time 3D Tracking Using Online and Offline Information, IEEE
Transactions on Pattern Analysis and Machine Intelligence, Vol. 26,
Nr. 10, pp. 1385-1391, 2004.
L. Vacchetti, V. Lepetit and P. Fua, Combining
Edge and Texture Information for Real-Time Accurate 3D Camera
Tracking, International Symposium on Mixed and Augmented Reality,
Arlington, VA, November 2004.
L. Vacchetti, V. Lepetit and P. Fua, Stable
3--D Tracking in Real-Time Using Integrated Context Information,
Conference on Computer Vision and Pattern Recognition, Madison, WI,
June 2003.
V. Lepetit, L. Vacchetti, D. Thalmann and P.
Fua, Fully
Automated and Stable Registration for Augmented Reality
Applications, International Symposium on Mixed and Augmented
Reality, Tokyo, Japan, 2003.
Contact: V. Lepetit vincent.lepetit@epfl.ch
|
