The invention relates to the technical field of
multimedia communication control, in particular to a method and
system device for controlling audio and video interaction in a multi-person conference, which comprises the following steps of: acquiring an active sound source
direction angle value in an audio channel, analyzing a change trend of a direction difference value, extracting a
delay value of channel audio and image loading time, and calculating a
delay value of the channel audio and image loading time; and identifying the fluctuation trend of the
barycentric coordinates of the face area, evaluating the position stability of the image main body, setting the user corresponding to the active channel as a main speaker role, and generating a main speaker
conference control authority
list. According to the invention, the audio activity and the
time data processing of the
image frame are combined, the synchronization between the audio and the video is optimized, the loading
delay is reduced, the smooth interaction of the conference is ensured, and the stability of the
barycentric coordinates of the face area in the image is analyzed, so that the
visual angle with the stable image main
body position can be effectively evaluated and selected; therefore, the front view can reflect the focus of the conference more accurately, and role distribution and authority management of the speaker are more intuitive and effective.