Investigation Of The Scalable Video Coding And Transport Control Based On Fish-Eye Bit Dropping Technique

With the development of new technique on Internet and customer's enthusiastic needs for network entertainment, video service will play an important role in the IP network without fail. It's really a challenge to apply the video service smoothly and reliably on the existing IP networks that had their geneses in an era when voice was the dominant service supported.As its prosperous potential, scalable video coding is becoming a hot topic in video transport. However, we find that traditional research on scalable video coding concentrates more on technology itself in network or coding/decoding than on subjective visual quality of the basic layer. In order to further develop the scalable video coding, some scholars are more concerned with the application of object-orentied scalable video coding. Due to the immature and complicated video object segmentation, the object-orentied scalable video coding is hard to implement in the near future. In this thesis, the scalable video coding based on fish-eye bit dropping technique which is supposed to fill up the gap is proposed.It is proved that fish-eye bit dropping shall achieve excellent performance in subjective visual feature by the experiment in this thesis. And this thesis confirmed further the compression video coding based on fish-eye bit dropping experiment under H.263 and H.264/AVC standard by using Visual C++. The experiment indicates that it has a high-quality for subjective visual sense under high compression ratios. Moreover, this thesis takes the research above as its foundation, and the scalable video coding and transport control based on fish-eye bit dropping technique is designed for H.263 and H.264 standard.As an improvement on traditional scalable video coding technique, the scalable video coding and transport control based on fish-eye bit dropping technique effectively enhance the subjective quality in lower bit rate condition. It provides a simple and practical means for the IP networks originally designed for data and voice services to support the video adaptive transportation more smoothly and reliably, thus it has the realistic significance.