Dynamic bit allocation in video combining for multipointconferencing

摘要

For multipoint video conferencing over a wide-area network, anvideo bridge is often used to combine the coded video signals fromnmultiple participants into a single video for display. The videoncombining can be done in the coded-domain by concatenating coded videonbitstreams, or in the pixel-domain by transcoding. In this paper wencompare the video quality using the two approaches for a multipointnvideo conference over a symmetrical wide-area network such as IntegratednService Digital Network (ISDN). We observe that in most multipoint videonconferences usually only one or two persons are active at one time,nwhile other participants are just listening with little motion. Tonachieve similar video quality for all the conference participants, thenactive persons require higher bit rates than the inactive persons. Byndoing video combining using the transcoding approach and dynamicallynallocating more bits to the active persons, we show that the achievednvideo quality can be much better than that can be achieved using thencoded-domain video combining approach