论文部分内容阅读
EPZS是联合多视点视频编码(JMVC,Joint Multi-view Video Coding)运动估计中采用的一种预测搜索算法,其搜索速度慢.针对EPZS算法的性能不足,我们在预测矢量集合、搜索模型、阈值设置和搜索策略四个方面进行改进,提出了一种混合快速搜索算法.在联合多视点视频编码测试平台JMVC8.3中,对三个由平行摄像机采集的多视点视频测试序列BallRoom、Exit和Vassar进行测试.实验结果表明:在保证视频重建质量和码率的前提下,与Jmvc中的EPZS算法相比,编码速度平均提高了55.66%～69.62%,改进算法的效果明显,编码效率得以提高.
EPZS is a kind of predictive search algorithm used in motion estimation of Joint Multi-view Video Coding (JMVC), and its search speed is slow.For the lack of performance of EPZS algorithm, we use a set of prediction vectors, search model, threshold Setting and searching strategies, a hybrid fast search algorithm is proposed.Under the joint multi-view video coding test platform JMVC8.3, three multi-view video test sequences BallRoom, Exit and Vassar The experimental results show that, compared with the EPZS algorithm in Jmvc, the encoding speed is improved by 55.66% ~ 69.62% on average, while the video reconstruction quality and bit rate are guaranteed. The improved algorithm has obvious effect and the coding efficiency is improved.