| 摘要 |
A method of providing, over a network, data to recreate an image in a device comprising: detecting the position of a foreground object in the original image and generating position information on dependence thereon; removing the foreground object from the image; and transferring to the device a background image comprising the original image with the foreground object removed, the removed foreground object and the position information. The background image may be scaled, divided into segments and coded. An area of the background image located around the foreground items of interest may be encoded at a higher bitrate. A depth map may also be streamed so that a stereoscopic image may be formed. A supplemental area in one of the segments may be used to transmit the removed foreground object. The method may be used to generate 3D images from 2D images and reduce the required bandwidth. |
