多视点视频编码预测结构的研究

Research on structure prediction of multiview video coding

不同的多视点视频序列具有不同的时空相关性,MVC采用的分层B帧预测结构无法充分利用视频序列的属性。本文提出了一种能适应不同视频序列编码的方法,通过研究多视点视频图像中的时空相关性和视点间的相关性,适当地取消部分时间层的视点间预测并且对所有P视点的非关键帧均采用视点间预测和调整了I视点的位置。实验结果表明,本文方法在提高编码效率的同时使得预测结构具有更好的随机访问性能、更低的编码复杂度。

Muhiview video sequences with different temporal and spatial correlation,the hierarchical B pictures coding structure used by MVC is Unable to take advantage of the properties of video sequences. A method adapting to different video sequence coding method is proposed based on research of temporal and inter-view correlation of mulitview video picture, Cancel inter-view prediction of some temporal level and select the appropriate position of I-view,and inter-view predicion is used in non-anchor frames of P-view.Experimental results show that the proposed method while improving coding efficiency and random access ability ,simultaneously reducing computation complexity.