| Summary: | 碩士 === 國立中正大學 === 通訊工程學系 === 98 === With the dramatically development of digitalization, the video streaming technologies via internet attract lots of attention in academic and industry in recent years especially for video conference. However, a number of well-known communication software, such as MSN and Skype, only provide peer-to-peer and limited video communication. Therefore, the development of video conference system for multi-user is needed. Besides, the low-cost technology is always essential to the implementation. To solve these issues, modified Scalable Video Coding (SVC) favors economical and efficient multimedia communications.
There are several years after the final proposals of H.264/AVC and the extended standards. There are three scalable options, including temporal scalability, SNR scalability and spatial scalability. Due to the highly complicated architecture in SVC only SNR scalability (CGS) and spatial scalability are discussed in this thesis. To improve the encoding speed, this thesis adopts x264 to design a scalable video system, and furthermore establishes the H.264/AVC-SVC transcoder to transform several H.264/AVC coded streams into a single SVC stream. Finally, the experimental results show the frame-rates of the transcoder with SNR scalability and spatial scalability are 41.3 fps and 44.02 fps, respectively. Moreover, the requirements for real-time system are satisfied without taking the network conditions into account.
|
|---|
