摘要
本文提出了两种高效率的用于匹配两个三维点集的方法,第一种方法是对文献[1]所提出的方法的改进,但第一种方法需要知道三个匹配点,为了放宽条件,我们提出了第二种方法,第二种方法具有更好的适用性,它可以解决任意两个三维点集的匹配问题,这两种方法都是基于先分别对两个点集中的全部点进行排序,然后进行小范围的局部匹配的思想;假定两个点集中都有几个点,它们的计算复杂性是O(n log n),而其它的方法都是O(n^2)。在点集较大的情况下(大于20个点),这两种方法将获得快速而有效的结果,并且容易逼近总体上的最佳匹配。
Two efficient algorithms for matching two 3-D point sets extracted from ri-.gid objects were proposed. The first is an improvement over the algorithm described in [1] , which requires an initial partial match of three noncollinear sensed points. The second algorithm can match general 3-D point sets without such strong requirements. Both algorithms are based on the idea of sorting all the points in each set first, then making local matching on fewer points. The matched points are guaranteed to be globally consistent. The computational complexity of our algorithms is O(nlogn) instead of O(n2), where n is the number of points in each set.
出处
《计算机学报》
EI
CSCD
北大核心
1989年第9期674-681,共8页
Chinese Journal of Computers
