Investigating the Usage of Modern Parallel Computing Platforms for Enhancing MVC Encoder Performance

Multi-see Video Coding (MVC) institutionalization is a progressing exertion planning to expand H.264/AVC by creating novel devices streamlined for 3D and multiview video use cases. One of the key recognized necessities for the MVC standard is its capacity to help parallel handling of various perspectives. The parallel task is known to be particularly significant for 3DTV applications, where the presentation needs to yield numerous perspectives at the same time to help head-movement parallax. we present a novel coding structure that empowers parallel encoder/decoder activity for various perspectives, without bargaining from the coding productivity. This is accomplished by methodicallly confining the reference zone of each view, with the goal that encoding and interpreting of large scale hinders from various perspectives could be proficiently pipelined and parallel task of independent perspectives ends up conceivable. As the between view forecast is as yet utilized, proposed structure accomplishes up to 0.9 dB addition contrasted with simulcast, keeping up fundamentally the same as alluring parallelism attributes.