Computes the geometric variance and number of valid correspondences between two point clouds. More...

Functions
void RTABMAP_CORE_EXPORT	rtabmap::util3d::computeVarianceAndCorrespondences (const pcl::PointCloud< pcl::PointNormal >::ConstPtr &cloudA, const pcl::PointCloud< pcl::PointNormal >::ConstPtr &cloudB, double maxCorrespondenceDistance, double maxCorrespondenceAngle, double &variance, int &correspondencesOut, bool reciprocal)
	Compute with variance and correspondences of `pcl::PointNormal` point cloud type.

void RTABMAP_CORE_EXPORT	rtabmap::util3d::computeVarianceAndCorrespondences (const pcl::PointCloud< pcl::PointXYZINormal >::ConstPtr &cloudA, const pcl::PointCloud< pcl::PointXYZINormal >::ConstPtr &cloudB, double maxCorrespondenceDistance, double maxCorrespondenceAngle, double &variance, int &correspondencesOut, bool reciprocal)
	Compute with variance and correspondences of `pcl::PointXYZINormal` point cloud type.

void RTABMAP_CORE_EXPORT	rtabmap::util3d::computeVarianceAndCorrespondences (const pcl::PointCloud< pcl::PointXYZ >::ConstPtr &cloudA, const pcl::PointCloud< pcl::PointXYZ >::ConstPtr &cloudB, double maxCorrespondenceDistance, double &variance, int &correspondencesOut, bool reciprocal)
	Compute with variance and correspondences of `pcl::PointXYZ` point cloud type.

void RTABMAP_CORE_EXPORT	rtabmap::util3d::computeVarianceAndCorrespondences (const pcl::PointCloud< pcl::PointXYZI >::ConstPtr &cloudA, const pcl::PointCloud< pcl::PointXYZI >::ConstPtr &cloudB, double maxCorrespondenceDistance, double &variance, int &correspondencesOut, bool reciprocal)
	Compute with variance and correspondences of `pcl::PointXYZI` point cloud type.

Detailed Description

Computes the geometric variance and number of valid correspondences between two point clouds.

This function estimates correspondences between two point clouds and computes a variance measure (based on squared median error) for the inlier correspondences. Optionally filters correspondences by normal alignment angle if the point cloud has normals.

The correspondence estimation uses PCL's reciprocal correspondence logic by default. The target and source clouds are automatically chosen based on their sizes (the larger becomes the target to ensure optimal matching behavior).

Parameters

	cloudA	First input point cloud (used interchangeably with cloudB for matching).
	cloudB	Second input point cloud.
	maxCorrespondenceDistance	Maximum allowable Euclidean distance between correspondences.
	maxCorrespondenceAngle	Maximum angle (in radians) between normals for correspondences to be considered valid. If ≤ 0, angle filtering is disabled.
[out]	variance	Output variance value (estimated using 2.1981 × median squared distance of inliers).
[out]	correspondencesOut	Output number of correspondences that passed all filters (distance and optional angle).
	reciprocal	If true, use reciprocal correspondences.

Note: This function chooses the target/source roles based on cloud size (the larger becomes the target).; The normal comparison is only applied if maxCorrespondenceAngle > 0.; The computed variance is based on a robust estimator using the median squared distance of correspondences.

See also: pcl::registration::CorrespondenceEstimation

Functions

Detailed Description