FPFH特征估计与关键点选择-优快云博客

本文链接：https://blog.youkuaiyun.com/Angle_best/article/details/53513883

本文探讨了PCL库中FPFH特征估计方法的使用细节，特别是当输入数据集为pcl::PointNormal类型时的可能性及效果。此外，还讨论了如何选择合适的特征关键点以提高FPFH特征的质量，并推荐了几种适用于FPFH的关键点检测算法。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

> Hello everyone,

> pcl::PointNormal contains informationabout the normal vector, the curvature and the xyz data of a point.

> Is it possible to pass to a FPFHfeature estimator the following input dataset?:

>pcl::FPFHEstimationOMP<pcl::PointNormal,pcl::PointNormal,pcl::FPFHSignature33> fpfhEst; instead of pcl::FPFHEstimationOMP<pcl::PointXYZ,pcl::Normal,pcl::FPFHSignature33> fpfhEst?

Yes, but only XYZ will be exploited for thefirst template and only

normal data will be used from the second.

> Will it still result in a good result?

It won't affect the results, you can giveit a try.

> What are good keypoints for FPFHfeatures?

> good FPFH features are calculated atpoints with a high curvature, so it would be good to calculate AGASTKeypoints whose are the corners in the frame, am i right?

You can calculate FPFH for any keypointsince they are disjoint processes. FPFH is a statistical descriptorthat computes histograms so question if Agast points are good keypointscan only be answered empirically since there is no directrelationship between the vicinity information (used in FPFH) and theintensity information (used in AGAST).I would rather suggestHarrisKeypoint3D which rely on geometric variation of the neighborhood to detectkeypoints the more variation you have the more your accumulated histogramwill look sharper and unique.

> Thanks for any help,