已知有效点数的点特征松弛匹配

Point Pattern Relaxation Matching with Known Number of Effective Points

研究了已知特征点集合中有效点数的先验知识时,对基本点特征松弛匹配算法的改进,包括松弛过程中初始匹配度量值的计算公式与松弛迭代公式,及确定对应点对的技术,以提高匹配性能。采用针对模拟图像及实际卫片航片图像对的实验,比较了两种方法的性能,验证了改进算法的正确性和有效性。

In case that the number of effective feature points is known, the original point pattern relaxation-matching algorithm is improved, including the computation of the initial support (merit value) and the iterative equation of relaxation matching. A better matching performance is obtained. Experimental results with large simulated images are given to compare the performance of the original method and algorithm proposed in this paper. The correctness and validity of the improved algorithm are verified.