Video compression based on distributed source coding principles

This paper applies distributed source coding principles to three areas of video compression, leading to Witsenhausen-Wyner video coding, layered Wyner-Ziv video coding, and multiterminal video coding. Compared to H.264/AVC and H.26L-FGS video coding, the advantage of Witsenhausen-Wyner video coding and layered Wyner-Ziv video coding is error robustness when the compressed video has to be delivered over a noisy channel. The price for this added error robustness is a small loss of compression performance. In multiterminal video coding for video sensor network applications, since the encoders are not allowed to communicate with each other, the use of distributed source coding is out of necessity. We propose a generic scheme for multiterminal video coding based on Slepian-Wolf coded quantization, where we explore multiple view geometry in computer vision to generate the side information. Experimental results with two- and three-view video sequences show that our multiterminal video coder can achieve a lower sum-rate than separate H.264/AVC coding. Comparison with the standard JMVM joint encoding scheme is also provided.