摘要
针对视频转码中运动矢量重估计精度不高的问题,提出了一种运动矢量自适应多模式精炼(AMPR)算法.该算法首先利用输入码流中运动矢量的相位和幅值信息构建了视频局部区域活动性模型,通过此模型自适应确定精炼窗口的大小,然后在菱形搜索算法(DS)和水平垂直搜索算法(HAVS)的基础上,给出了多模式搜索策略.仿真实验表明,该算法取得接近于全搜索算法的视频质量,并有效地降低了计算复杂度.
An adaptive multi-pattern refinement algorithm for motion vector in video transcoding is presented. Using the consistency of angle and amplitude of intensity of local motion vectors, a motion activity model is constructed, so the search range of the rnacroblock is confirmed adaptively using the activity model. Then, a multi-pattern refinement strategy based on the diamond search and horizontal and vertical search is proposed. Experimental results show that the proposed method reduces the computation complexity significantly without degrading the quality of video transcoder.
出处
《微电子学与计算机》
CSCD
北大核心
2010年第9期24-28,共5页
Microelectronics & Computer
基金
国家自然科学基金项目(60602034)
关键词
视频转码
运动判定
空间下采样
自适应精炼
video transcoding
motion estimation
space downscaling
adaptive refinement
