一种块匹配的快速运动估计算法

A Fast Block-matching Motion Estimation Algorithm

为了减小运动估计算法的计算复杂度及提高视频压缩编码的效率,提出了一种新的快速、有效的块匹配运动估计算法。该算法首先采用多步搜索策略,并利用相邻块的运动矢量来预测当前块的搜索起点,然后用梯度下降的方向来指导搜索方向,以节约计算资源,对于最佳匹配点所在区域,则采用菱形与正方形相结合的模板来进行精细搜索,以提高算法的搜索精度。实验结果证明,该算法在降低计算复杂度的同时,还能有效地保证搜索精度。

A new fast and efficient search algorithm for block motion estimation is presented in this paper, The new algorithm can reduce the computational complexity of the fast motion estimation and improve the efficiency of video compression code. The algorithm accepts a multi-step search strategy, using motion vectors of the adjacent block to predict starting search point, using gradient descent method to optimize the search to reduce computation resource. For the area where optimal points located, a Diamond-Square Mask was used to carry out the refined search. Experimental results show that the proposed algorithm can reduce computational complexity efficiently, and at the same time while, ensuring the search accuracy.