序列图像差分关联域的复杂全域运动估计

Complicated global motion estimation based on sequence image difference correlative block

本文提出了一种新的基于差分关联域的复杂全域运动估计算法,该算法通过选取图像中梯度大的区域,避免了块匹配算法的误匹配问题,使算法的准确性得到保障。而且该算法结合运动估计技术中的参数估计与非参数估计技术,扩展了块匹配算法的适用范围,使得对复杂的全域运动进行有效估计成为可能。理论分析及大量实验证明,此算法能有效地处理包含缩放、旋转等分量的复杂运动。而且本算法将复杂的全域运动分解为各基本运动,使系统可以根据运算量的限制自适应地选择最优的运动估计算法。

A new motion estimation algorithm on difference correlative block was introduced. The algorithm avoided the mistake of matching by choosing the area of big gradient in the images to ensure the precision. And, this algorithm integrated the parameter and non-parameter estimation to extend the applicable range of block matching, and made the effective estimation of complicated global motion possible. The theoretical analysis and experimental results demonstrate that this algorithm can deal with complicated global motion effectively, including zoom, rotation motion. Furthermore, global motion is disassembled into basis motion, and the adaptive system can choose optimal algorithm according to limitation of operation burden.