Research And Implementation Of Network Virtual Video Karaoke Synthesis System Based On Directshow

With the development of digital media technology and network technology, the research of Augmented Reality based on the integration of motion video has become a key research project of the related fields of computer graphics and computer vision, which is also the trend of the development of digital media technology. This technique is often able to integrate specific digital scenes with real shooting video efficiently and timely, which gives people immersive feelings and entertainment experiences.This article is based on augmented reality technology of the dynamic video integration, which develop a new entertainment system------network virtual video karaoke synthesis system. The system intends to position singers who are at two different nodes on the network into the same virtual stage by keying integration technology, visualizes the virtual chorus effect on the singers’ respective terminals timely, also provide humane functions such as local storage, recording and collection, and ultimately makes the people in different regions to have the feeling of "Although far away, just around the corner".The system is structured by using the currently popular DirectShow framework in the streaming media processing, and each module is packaged a compatible DirectShow Filter module. Among them, keying fusion Filter is achieved by using background difference keying algorithm and the OpenCV library; network video-audio sender and receiver Filter is achieved by using WinSock technology. Finally proved by experiments, the system has the feature of good keying integration effect and simple operation, which basically achieves the desired objectives.