一种有效的三步运动估计算法

An Effective Three-step Search Algorithm for Motion Estimation

为了减小运动估计算法的计算复杂度和提高红外图像超分辨率重建的可靠性,提出了一种有效的三步搜索算法。该算法采用多步搜索策略,根据运动矢量分布的中心偏移性及并行处理的思想,在最佳匹配点所在的区域使用菱形小模板代替新三步法中的正方形小模板来进行精细搜索,以提高算法的搜索精度。实验结果表明,与全搜索法、菱形搜索算法和预测性菱形搜索算法相比,该算法能在保证搜索精度的同时大大缩短消耗时间。

To reduce the computation complexity of motion estimation algorithms and improve the reliability of super-resolution reconstruction of infrared images, an effective three-step search algorithm is proposed. In the algorithm, a multi-step search strategy is adopted. According to the center offset characteristics of a motion vector and the idea of parallel processing, it makes a fine search in the region in which the best matching points exist by using a small diamond search template instead of a small square template. Thus, its search accuracy is improved. The experimental result shows that compared with other search algorithms, this search algorithm can reduce the calculation time greatly while its search accuracy is ensured.