Virtual Post-its: Visual Label Extraction, Attachment, and Tracking for Teleconferencing

We present a novel functionality for tele-conferencing and tele-teaching, coined 'virtual post-IT'. A virtual label can be cut out of a scene or can have been determined beforehand. This label can then be virtually attached to either a part of the screen or to an object in the scene. When made part of the screen, it remains there as a reminder, e.g. to a complicated formula. When glued to an object, the label automatically follows it when the speaker moves the object around. This real-time tracking includes out-of-plane rotation and does not require any special markers. The selection of the pattern to be used as a label can be carried out on-line by the speaker, by simply encircling it with a laser pointer. S/he can also determine on which object it has to be glued and where. Also these commands are given through simple, intuitive motions with a laser pointer. As the encircling motion to cut out the label will be rather sloppy, the system autonomously determines a rectangle that encloses the data and that has edges which are aligned with the dominant directions of the environment (typically horizontal and vertical).

[1]  John K. Tsotsos,et al.  SAVI: an actively controlled teleconferencing system , 2001, Image Vis. Comput..

[2]  Luc Van Gool,et al.  Content-Based Image Retrieval Based on Local Affinely Invariant Regions , 1999, VISUAL.

[3]  Milton Chen,et al.  Design of a virtual auditorium , 2001, MULTIMEDIA '01.

[4]  Luc Van Gool,et al.  Markerless augmented reality with a real-time affine region tracker , 2001, Proceedings IEEE and ACM International Symposium on Augmented Reality.

[5]  Luc Van Gool,et al.  Local Features for Image Retrieval , 1999, State-of-the-Art in Content-Based Image and Video Retrieval.

[6]  A. Raczynski,et al.  Services and training through augmented reality , 2004 .

[7]  Andrew Zisserman,et al.  Multiple view geometry in computer visiond , 2001 .