interface for the implement of basic icp algorithm using pcl

pertubate the point cloud with a small translation

Brief: Get rid of some unstable points from the keypoints list

point-to-line LLS (calculated using primary direction vector), used now the normal vector here actually stores the primary direcyion vector (for easie

primary vector stored as point normal

point-to-line LLS (calculated using normal vector) , Deprecated

ground 3dof : roll, pitch, z

point-to-plane LLS

point-to-point LLS

fixed number random downsampling when keep_number == 0, the output point cloud would be empty

Brief: Read the point cloud data of various format

Used for coordinate system convertation

read the coordinate of the line begining with target_station_name format: station_name x y z

TODO:

filter the ghost points under ground and the points on the ego-vehicle ghost radius should be a bit larger than self radius we assume that most of the

\brief Search for the closest nearest neighbor of a given point. * \param query the point to search a nearest neighbour for * \param i

\brief Set the number of neighbors used when selecting a point neighbourhood * to compute covariances. * A higher value will bring mor

\brief Set the minimum allowable ratio between eigenvalues to prevent singular covariance matrices. * \param[in] min_covar_eigvalue_mult the m

\brief Provide a pointer to the input dataset * \param cloud the const boost shared pointer to a PointCloud message */

\brief Provide a pointer to the input target (e.g., the point cloud that we want to align the input source to). * \param[in] cloud the input point

\brief Provide a pointer to the input target (e.g., the point cloud that we want to align the input source to) * \param[in] target the input p

\brief Set the minimum number of points required for a cell to be used (must be 3 or greater for covariance calculation). * \param[in] min_poi

\brief Set/change the point cloud outlier ratio. * \param[in] outlier_ratio outlier ratio */

\brief Set/change the voxel grid resolution. * \param[in] resolution side length of voxels */

\brief Set the rotation epsilon (maximum allowable difference between two * consecutive rotations) in order for an optimization to be consider

\brief Provide a pointer to the covariances of the input source (if computed externally!). * If not set, GeneralizedIterativeClosestPoint will

\brief Set/change the newton line search maximum step length. * \param[in] step_size maximum step length */

\brief Provide a pointer to the covariances of the input target (if computed externally!). * If not set, GeneralizedIterativeClosestPoint will

G2O has some problem, use ceres instead

mapping from [0,1] to [0,1]

using uniform motion model to do motion compensation

using uniform motion model to do motion compensation (with time stamp (store timestamp as curvature))

Given a matrix M/A[1..m][1..n], this routine computes its singular value decomposition, M/A = U·W·V T. Thematrix U replaces a on output. The diagonal

Delete invalid characters such as spaces, tabs, etc. in strings

hard-coded (not a good way) to adjust the parameters on fly