H.264/AVC Scalable Video Coding (SVC) is an emerging video coding standard developed by the Joint Video Team
(JVT), which supports multiple scalability features. With scalabilities, SVC video data can be easily adapted to the
characteristics of heterogeneous networks and various devices. Furthermore, SVC requires a high coding efficiency that
is equally competitive or better than single-layer H.264/AVC. Motion prediction at the level of Fine Grain Scalability
(FGS) enhancement layers was proposed to improve coding efficiency as well as inter-layer motion prediction. However,
the removal of the FGS enhancement layer at the inter-layer motion prediction causes significant visual errors due to the
encoder-decoder mismatches of motion vectors and MB modes. In this paper, we analyze visual errors to find the cause
as well as the method for reducing such errors. Experimental results showed that the proposed method allowed SVC
bitstreams decoding with reduced visual errors, even when the FGS enhancement layer used for the inter-layer motion
prediction was removed.
For outstanding coding efficiency with scalability functions, SVC (Scalable Video Coding) is being standardized.
SVC can support spatial, temporal and SNR scalability and these scalabilities are useful to provide a smooth video
streaming service even in a time varying network such as a mobile environment. But current SVC is insufficient to
support dynamic video conversion with scalability, thereby the adaptation of bitrate to meet a fluctuating network
condition is limited. In this paper, we propose dynamic full-scalability conversion methods for QoS adaptive video
streaming in SVC. To accomplish full scalability dynamic conversion, we develop corresponding bitstream extraction,
encoding and decoding schemes. At the encoder, we insert the IDR NAL periodically to solve the problems of spatial
scalability conversion. At the extractor, we analyze the SVC bitstream to get the information which enable dynamic
extraction. Real time extraction is achieved by using this information. Finally, we develop the decoder so that it can
manage the changing scalability. Experimental results showed that dynamic full-scalability conversion was verified and
it was necessary for time varying network condition.