新的视频标准H.264为了获得更精确的运动向量和更高的压缩比,引入了分数像素运动补偿技术,但同时也增加了运动补偿过程的复杂度。为了克服这一局限,根据相邻分数像素点之间高度相关的运动估计匹配误差和运动矢量方向,以及运动向量具有中心偏置的特性提出了一种新的1/4像素精度快速搜索算法(QAFSA)。实验表明:该算法在保持图像质量和码率基本不变的情况下大大减少了搜索点的数目,有效地提高了分数像素运动估计的速度。
H. 264 is a new video standardization in which fractional pixel motion compensation is applied in order to achieve more accurate motion description and higher compression efficiency. However, it remarkably increases complexity in motion estimation process. Therefore reducing the computation load incurred by fractional pixel search is necessary. This paper proposed an efficient fast fractional pixel motion estimation algorithm (QAFSA) based on the highly correlation of motion estimation cost, motion vector direction and the center-blased of motion vector for H. 264. Test results show that the proposal algorithm greatly reduces the number of searching points and enhances the speed of motion estimation, while keeping the nearly equal quality of the image.