Scalable Video Coding Based on Three-Dimensional Discrete Pseudo Cosine Transform

One of the top tasks for video streaming, video surveillance, video conferencing and video broadcasting is a low bit rate and real-time video coding. The most popular low bit rate video compression approach is based on the H.264/AVC standard [1]. This standard provides high compression efficiency due to hybrid video coding which includes intra-frame prediction, variable-size block motion estimation and compensation, variable-size transform, high efficiency entropy coding etc. As an alternative, this presentation describes a low bit rate video coding approach based on threedimensional discrete pseudo cosine transform (3-D DPCT) [2]. This approach uses faster arithmetic coder implementation than H.264/AVC, does not use motion estimation for exploiting video source temporal redundancy and as well as H.264/AVC uses multiplication-free transform and divisionfree quantization. From this it follows that 3-D DPCT codec has lower computational complexity than main profile of H.264/AVC standard. At the same time it has high compression efficiency at low bit rates. Therefore, proposed video coding algorithm can be attractive for low bit rate video communication. In addition video source rate control based on adaptive Lagrange multiplier selection is introduced [3]. Practical results which show the compression efficiency of the proposed algorithm in comparison with H.264/AVC standard and its scalable (SVC) extension are presented.