基于改进的部分Hausdorff距离的图像配准

摘要

After obtaining the edge of reference image and benchmark image, the improved partial Hausdorff distance as image similarity measure registration is used. In the search strategy, a genetic algorithm is adopted to speed up the search. It not only quickens the matching process, improves the antinoise performance, and also accurately matches occluded image. A lot of experiments show that the algorithm is robust, the precision of matching is high and the computing speed is fast.%在获取基准图与参考图边缘的基础上,采用改进的部分Hausdorff距离作为相似性度量对图像进行配准.在搜索策略上,采用遗传算法加快了搜索的速度.该方法不仅加快了匹配过程,提高了抗噪性能,而且能准确匹配遮挡图像.实验证明,这种算法鲁棒性好、匹配精度高、计算速度快.