Methods and system for simulated 3D videoconferencing
A system and method for manipulating images in a videoconferencing session provides users with a 3-D-like view of one or more presented sites, without the need for 3-D equipment. A plurality of cameras may record a room at a transmitting endpoint, and the receiving endpoint may select one of the received video streams based upon a point of view of a conferee at the receiving endpoint. The conferee at the receiving endpoint will thus experience a 3-D-like view of the presented site.
1. A continuous presence videoconferencing apparatus, comprising:
a first plurality of cameras, arranged in an arc located in a horizontal plane above a video display to capture image data corresponding to a conferee during a video conference, each camera of the first plurality of cameras configured to create a video stream; and
an endpoint configured to compress each video stream created by the first plurality of cameras into a first single multi-view video stream, the endpoint further configured to receive a second multi-view video stream from a remote endpoint, the second multi-view video stream including a plurality of compressed video streams having originated at a second plurality of cameras at the remote endpoint;
a point-of-view detector configured to determine a gaze direction of the conferee relative to the video display based on the image data captured by one or more of the first plurality of cameras; and
an editor, communicatively coupled to the point-of-view detector, the editor configured to build an image responsive to the gaze direction of the conferee and cause the video display to display the image, wherein building an image responsive to the gaze direction of the conferee comprises selecting one of the compressed video streams within the second multi-view video stream received from the remote endpoint in correspondence to the gaze direction of the conferee relative to the video display.
2. The continuous presence videoconferencing apparatus of claim 1 , wherein the first plurality of cameras comprises at least three video cameras.
3. The continuous presence videoconferencing apparatus of claim 1 , further comprising a control module coupled to the point-of-view detector, the control module configured to design a layout and to provide information to the point-of-view detector on each segment of the layout.
4. The continuous presence videoconferencing apparatus of claim 1 , wherein the point-of-view detector is further configured to:
detect another gaze of the conferee toward the displayed image;
determine a portion of the displayed image to which the conferee looks; and
make a selection based, at least in part, on the determined portion.
5. The continuous presence videoconferencing apparatus of claim 1 , wherein the point-of-view detector is further configured to determine a gaze of the conferee as an angle relative to a line perpendicular to the display.
6. The continuous presence videoconferencing apparatus of claim 1 , wherein the point-of-view detector is further configured to determine a gaze of the conferee as a two-dimensional pixel count relative to a pixel at a predetermined position within the displayed image.
7. The continuous presence videoconferencing apparatus of claim 1 , wherein the first plurality of cameras includes at least one camera configured to respond to near infra-red light, and wherein the point-of-view detector is further configured to respond to near infra-red information received from the at least one camera to determine a gaze of the conferee.