3D Visual Communications

5.1 Fundamentals of Video Coding

With the advancement of 2D video compression technology, end users can experience better visual quality at much lower bit rate. To take advantage of the successful development from existing 2D video coding tools and their fast deployment in the marketplace, the 3D video codecs are often built based on the popular 2D video codecs, especially with consideration of the availability of hardware platforms. The mature state-of-the-art video codec, H.264 (MPEG-4 Part 10, Advanced Video Coding (AVC)) [1], has often been selected as a fundamental building block and code base to construct advanced 3D video codecs. Before we introduce the 3D video codecs, we first overview the fundamentals of 2D video coding and the important components adopted in H.264. A short overview of the next generation video codec, high efficiency video coding (HEVC), will be presented at the end of this section.

The modern consumer video codecs often deploy lossy compression methods to achieve higher compression efficiency, so the reconstructed pixels in each frame may not be bit-to-bit accurate when compared to the original video data. The degradation of the perceived video quality is related to the encoded bit rate: a stream with higher bit rate can provide better visual quality and vice versa. The key methodologies adopted in lossy video compression are prediction, transform, quantization, and entropy coding technologies. The concept of a prediction process consists of two coding ...