基于特征点检测的CPD配准算法

CPD Algorithm for Registration Based on Feature Point Detection

文章针对非刚性配准CPD(coherent point drift)算法的过程中,计算复杂运行时间长的问题,提出了一种基于曲率特征点检测的配准算法。通过提取目标点集的特征点集,并对所提取的特征点集用CPD算法进行配准,可以使减少运算规模进而减少算法运行时间,提高了配准效率。

In order to solve the problem of complex calculation and long running time in the process of non-rigid registration of CPD（coherent point drift） algorithm, a registration algorithm based on curvature feature point detection is proposed. By extracting the feature points of the target point set and registering the extracted feature points with the CPD algorithm, the operation can be less complex and the algorithm running time can be reduced, thereby improving the registration efficiency.