An object-oriented cognitive source coding architecture for 3D video communications

Reliable transmission of 3D video signals is nowadays an interesting research issue both for the new coding challenges that three-dimensional video signals pose and for the wide diffusion of multimedia communications over wireless networks. In order to deal effectively with packet losses over radio channels, several robust source coding schemes have been proposed. This article presents a reconfigurable and flexible architecture (named Cognitive Source Coder in analogy with Cognitive Radio systems) that implements different robust source coding solutions and adaptively adopts them according to channel conditions. The proposed approach permits improving the quality of the 3D scene reconstructed at the end terminal with respect to the corresponding non-adaptive approaches.