机构地区: 宁波大学信息科学与工程学院
出 处: 《电路与系统学报》 2009年第2期111-117,110,共8页
摘 要: 本文分析了多视点视频编码中宏块模式选择情况的统计特性,把宏块模式分为四类。提出了基于宏块模式类率失真均值的多阈值快速宏块模式选择算法及阈值的近似计算方法和动态更新算法,理论分析了该快速算法的编码率失真性能。实验结果表明,所提出快速算法在几乎不影响编码率失真性能的情况下,编码速度比多视点视频编码校验模型提高1.92倍到7.07倍。 After analyzing the statistical features of multiview video coding and dividing the Macroblock modes into 4 classes, the paper first proposes a multi-threshold fast macroblock selection algorithm based on the average values of rate and distortion of macroblock mode class, and then puts forward an approximate computing method of the multi-thresholds and an updating algorithm. Finally, the paper theoretically analyzes the rate and distortion performance of the fast algorithm. Experiment results show that the proposed fast algorithm promotes the encoding speed by 1.92-7.07 times in comparison with JMVM, while the algorithm hardly influences the rate and distortion performance.