Description:
In this video, the motion of a human is captured. The system automatically initializes itself. Throughout the tracking, skin color segmentation is used to track the hands and head. The shirt color is used for foreground segmentation. The output of the system is first demonstrated by the projection of the prediction of the human model, and then by applying the predicted configurations to a 3D visulization of the human model.

Player Information:
The video is encoded with mpeg4. For example, the VLC player plays the video.

Contact Information:
Fur further information, please send an email to Pedram Azad <azad@ira.uka.de>.


Attachment filename: 1134_VI.mpg (3714990 Bytes)