基于K近邻局部结构的截断二次谱匹配算法

摘要

针对存在大量非对应点的点模式匹配问题，提出了一种基于 K 近邻局部结构的截断二次谱匹配算法。该算法首先对点模式利用截断模式谱方法匹配得到初步匹配结果。再以匹配点对为参考点构造 K 近邻局部结构，对局部结构利用谱方法进行二次匹配来寻找剩下的匹配点。两次匹配的点对合并得到最终的匹配结果。仿真结果表明该方法在不同条件下都具有较好的性能。%Aiming at the problems of point pattern matching with large ammounts of non-corresponding points,a truncated secondary spectral matching method based on the K neighboring local structure is proposed. Truncated spectrum matching is used firstly to get the preliminary result,and the K neighboring local structure is constructed by matching point pairs.The spectrum method is used again in the local structure to find the rest of the matching points.The matching points are merged as the final result.Simulation results show that the proposed algorithm has a good performance in different conditions.