This can be used to synthesize virtual views different from the acquired ones. Multi-view video (MVV) is one of the key elements of these applications it consists in the simultaneous representation of a scene captured by N cameras placed in different spatial positions, called points of view.īy using more than two cameras during video acquisition, adjacent views act like local stereo pairs to guarantee stereoscopy to the viewer.
3D cinema productions have already generated big revenues, but other applications such as 3DTV and Free Viewpoint TV (FTV) are also becoming more desirable due to the increased affordability of 3D displays for home use. While traditional 2D video offers the viewer only a passive view of the scene, a more realistic experience can be obtained through applications such as 3D video or free viewpoint selection. In recent years, the advances in video acquisition, compression, transmission, and rendering have made possible the development of technologies that can enhance the viewers’ experience by including the third dimension. Our experimental results prove that our preference-based approach provides a high-quality decoding even when the uplink capacity of each node is only a small fraction of the rate of the stream. Using network coding enables the users to retrieve the content in a faster and more reliable manner and without the need for coordination among the senders. We propose a novel method for selection and network encoding of the transmitted frames based on the users’ preferences for the different views and the rate-distortion properties of the stream. In this paper, we address the topic of cooperative streaming of multi-view video content, wherein users who recently acquired the content can contribute parts of it to their neighbors by providing linear combinations of the video packets. Because of the high bandwidth cost they come with, multi-view streaming applications can greatly benefit from the use of network coding, in particular in transmission scenarios such as wireless network, where the channels have limited capacity and are affected by losses. Multi-view video streaming is an emerging video paradigm that enables new interactive services, such as 3D video, free viewpoint television, and immersive teleconferencing.