Real-time SVC bitstream extraction for a dynamic adaptation in video streaming

Abstract

The video streaming service is widely used thanks to the development of network environment. The scalable video coding is the coding scheme that supports the spatial, temporal, and quality scalability under being standardization by JVT. As using SVC bitstream, it is possible to serve dynamically adapted video streaming by extracting SVC bitstream with changed scalability according to current network condition. It is very important to find how the extractor extracts the necessary data from the SVC bitstream extracts in real-time. In this paper, we propose the real-time SVC bitstream extractor for dynamic adaptation in video streaming under dynamically changing network environment. First, we check the constraints for proposing extractor. And then, we propose the structure of real-time SVC bitstream extractor which consists of two parts: bitstream analyzer and bitstream extractor. And we propose the algorithm for extraction process of proposed extractor. Lastly, we propose the extraction methods of the bitstream with varied scalability for spatial, temporal, and quality scalability. We implemented the proposed extractor by using JSVM 2.0. To check validity of the proposed extraction method and extractor, we test the performance of proposed extractor by comparing the required bitrate with bitrate of extracted bitstream. From the experiments, we could verify that the proposed real-time SVC bitstream extractor extracts the bitstream with changed scalability according to varying the input parameters of the proposed extractor