wish list concept

[cwmeijer]

based on:
http://www.pointclouds.org/assets/iros2011/registration.pdf

Refining initial alignments using ICP:

• pcl::IterativeClosestPoint<PointT, PointT> icp;

// Threshold for outliers. If to far away between matches, ignore it while calculating rms.
? icp.setMaxCorrespondenceDistance (distance)

? icp.setRANSACOutlierRejectionThreshold (distance);

// ICP stops when the epsilon (difference) between the previous transformation and the current estimated transformation is smaller than an user imposed value (via setTransformationEpsilon)

• icp.setTransformationEpsilon (transformation_epsilon);

// ICP stops after n iterations.

• icp.setMaximumIterations (max_iterations);

// ICP stops when error between correspondences is smaller than this threshold.

• icp.setEuclideanFitnessEpsilon(distance)

// Provide a pointer to the input source (e.g., the point cloud that we want to align to the target)

• icp.setInputCloud (aligned_source);

// Provide a pointer to the input target (e.g., the point cloud that we want to align to the target)

• icp.setInputTarget (target_points);
• virtual void pcl::IterativeClosestPoint< PointSource, PointTarget, Scalar >::setInputTarget ( const PointCloudTargetConstPtr & cloud )
• PointCloud registration_output;

// Call the registration algorithm which estimates the transformation and returns the transformed source (input) as output.

• icp.align (registration_output);
• void pcl::Registration< PointSource, PointTarget, Scalar >::align ( PointCloudSource & output )

Get the final transformation matrix estimated by the registration method

• Eigen::Matrix4f refined_T = 14 icp.getFinalTransformation () * initial_T);
• Matrix4 pcl::Registration< PointSource, PointTarget, Scalar >::getFinalTransformation ( )

[larsmans]

I put a ? before icp.setMaxCorrespondenceDistance as well, since it's not clear to me if it really does outlier filtering at all.

[larsmans]

Preliminary wrapper for ICP in b4970c5 (branch registration).