A modern conferencing system has, next to voting and audio, also support for video.
This video is typically recorded by multiple wall-mounted cameras, programmed to show however is speaking.
The delegates (the participants of the meeting) can see this video stream on their meeting device (called the delegate unit).
This thesis researches the possibility to add annotation (e.g. the name of who is speaking) on the video, while maintaining ultra-low latency requirements. Possible extensions are: transcoding to different video standards; integrating with data sources such as twitter to annotate the video; …
This thesis consists of multiple phases:
Get aqcuinted with the video transport protocol used by Televic Conference
Literature study of real-time video processing libraries