基于运动矢量对消和差分原理的快速全局运动估计

Fast Global Motion Estimation Based on Symmetry Elimination and Difference of Motion Vectors

该文针对全局运动估计计算复杂度过高的问题,提出了一种基于运动矢量对消和差分原理的快速全局运动估计方法。基于不同象限运动矢量对之间存在的对称抵消特性,首先估计出平移运动参数分量;再使用运动矢量对的差分原理,并结合一种置信判断的策略,估计出变换运动参数分量。在5组几何全局运动模型和真实视频序列上的实验结果表明,与基于迭代最小二乘(ILS)和基于偏导数(PD)的估计算法相比,该方法在保证估计精度的情况下,所需的平均计算时间只要它们的50%左右。

To reduce the computational complexity of Global Motion Estimation （GME）, a fast GME method based on the principle of the symmetry elimination and difference of motion vectors is proposed. First, the translational parameters are estimated by using the technique of the symmetry elimination of motion vectors. And then the transform parameters are estimated by utilizing the principle of the difference of motion vectors and the strategy of the belief judgment. Experimental results on five geometric global motion models and the real video sequence show that the proposed method, compared with the estimation methods based on the Iterative Least Square （ILS） and the Partial Derivative （PD）, only need approximately 50% of their computational time while achieving the comparable estimation accuracy.