Visual Media Coding and Transmission

4.2 Distributed Source Coding

DVC is identified as the adaptation of the theoretical framework of distributed source coding (DSC) set by the Slepian–Wolf theorem [1] and the Wyner–Ziv theorem [2] for video coding. DSC concept deviates from the conventional source coding paradigm in the context of the dependency of the encoding of statistically-correlated sources. In the conventional approach, the statistically-correlated sources are jointly encoded and jointly decoded for the perfect reconstruction of the information stream at the decoder. DSC, in contrast, proposes to carry out independent encoding of statistically-dependent sources, and yet to jointly decode them. The information theoretical limitations inherited by the so-called independent encoding have been reported by Slepian and Wolf, as discussed in the next section.

4.2.1 The Slepian–Wolf Theorem

(Portions reprinted, with permission, from M.B. Badem, W.A.R.J. Weerakkody, W.A.C. Fernando, A.M. Kondoz, “Design of a non-linear quantizer for transform domain DVC”, IEICE Transactions on Fundamentals. Paper Number 2008EAP1119. ©2008 IEICE.)

Assume that X and Y are two statistically-dependent discrete random sequences which are independently and identically distributed (i.i.d.). Consider the case that these sequences are separately encoded with rates R_X and R_Y, respectively, but are jointly decoded, exploiting the correlation between them as illustrated in Figure 4.6. Slepian and Wolf have presented an analysis of the possible ...