A new Scalable Multi-View Video Coding configuration for mobile applications

Transmission of multi-view video content is not practical in most mobile environments due to the limited bandwidth and processing power of mobile devices. To support such environments, one can limit the number of views that are being transmitted, known as Scalable Multi-view Video Coding (SMVC). In this paper, we propose a new view selection method for view scalability in multi-view video coding in mobile environments, which uses inter and intra view dissimilarities to determine the most suitable views for the base layer corresponding to the prediction structure and user selected limited number of views. By selecting more correlated views for the base layer, the proposed method provides an improved performance, as confirmed by simulation results, even when all the enhancement layers are dropped due to network limitations.