Class NDT

Nested Relationships

Nested Types

Inheritance Relationships

Base Types

  • public mrpt::maps::CMetricMap

  • public mrpt::maps::NearestNeighborsCapable

  • public mp2p_icp::NearestPlaneCapable

Class Documentation

class NDT : public mrpt::maps::CMetricMap, public mrpt::maps::NearestNeighborsCapable, public mp2p_icp::NearestPlaneCapable

3D NDT maps: Voxels with points approximated by Gaussians. Implementation of magnusson2007scan.

This implements two nearest-neighbor virtual APIs:

  • mrpt::maps::NearestNeighborsCapable: It will return pairings of local points against global (this map) points in voxels that do not conform planes.

  • mp2p_icp::NearestPlaneCapable: It will return points-to-plane pairings, if the local point falls within a voxel with points conforming a plane.

Indices and coordinates

using global_plain_index_t = uint64_t

collapsed plain unique ID for global indices

static constexpr size_t MAX_POINTS_PER_VOXEL = 16
static constexpr std::size_t GLOBAL_ID_SUBVOXEL_BITCOUNT = 5
inline global_index3d_t coordToGlobalIdx(const mrpt::math::TPoint3Df &pt) const
inline mrpt::math::TPoint3Df globalIdxToCoord(const global_index3d_t idx) const

returns the coordinate of the voxel “bottom” corner

static inline global_plain_index_t g2plain(const global_index3d_t &g, int subVoxelIndex = 0)

Data structures

using grids_map_t = tsl::robin_map<global_index3d_t, VoxelData, index3d_hash<int32_t>>
bool ndt_is_plane(const mp2p_icp::PointCloudEigen &ndt) const

Returns true if the fitted Gaussian fulfills the requirements of being planar enough, according to the insertionOptions parameters

Basic API for construction and main parameters

NDT(float voxel_size = 5.00f)

Constructor / default ctor.

Parameters:

voxel_size – Voxel size [meters]

~NDT()
void setVoxelProperties(float voxel_size)

Reset the main voxel parameters, and clears all current map contents

Data access API

inline VoxelData *voxelByGlobalIdxs(const global_index3d_t &idx, bool createIfNew)

returns the voxeldata by global index coordinates, creating it or not if not found depending on createIfNew. Returns nullptr if not found and createIfNew is false

Function defined in the header file so compilers can optimize for literals “createIfNew”

inline const VoxelData *voxelByGlobalIdxs(const global_index3d_t &idx) const
inline VoxelData *voxelByCoords(const mrpt::math::TPoint3Df &pt, bool createIfNew)

Get a voxeldata by (x,y,z) coordinates, creating the voxel if needed.

inline const VoxelData *voxelByCoords(const mrpt::math::TPoint3Df &pt) const
void insertPoint(const mrpt::math::TPoint3Df &pt, const mrpt::math::TPoint3Df &sensorPose)

Insert one point into the map. The sensor pose is used to estimate the outward surface normal

inline const grids_map_t &voxels() const
mrpt::math::TBoundingBoxf boundingBox() const override

Computes the bounding box of all the points, or (0,0 ,0,0, 0,0) if there are no points. Results are cached unless the map is somehow modified to avoid repeated calculations.

void visitAllPoints(const std::function<void(const mrpt::math::TPoint3Df&)> &f) const
void visitAllVoxels(const std::function<void(const global_index3d_t&, const VoxelData&)> &f) const
bool saveToTextFile(const std::string &file) const

Save to a text file. Each line contains “X Y Z” point coordinates. Returns false if any error occured, true elsewere.

API of the NearestNeighborsCapable virtual interface

bool nn_has_indices_or_ids() const override
size_t nn_index_count() const override
bool nn_single_search(const mrpt::math::TPoint3Df &query, mrpt::math::TPoint3Df &result, float &out_dist_sqr, uint64_t &resultIndexOrID) const override
bool nn_single_search(const mrpt::math::TPoint2Df &query, mrpt::math::TPoint2Df &result, float &out_dist_sqr, uint64_t &resultIndexOrID) const override
void nn_multiple_search(const mrpt::math::TPoint3Df &query, const size_t N, std::vector<mrpt::math::TPoint3Df> &results, std::vector<float> &out_dists_sqr, std::vector<uint64_t> &resultIndicesOrIDs) const override
void nn_multiple_search(const mrpt::math::TPoint2Df &query, const size_t N, std::vector<mrpt::math::TPoint2Df> &results, std::vector<float> &out_dists_sqr, std::vector<uint64_t> &resultIndicesOrIDs) const override
void nn_radius_search(const mrpt::math::TPoint3Df &query, const float search_radius_sqr, std::vector<mrpt::math::TPoint3Df> &results, std::vector<float> &out_dists_sqr, std::vector<uint64_t> &resultIndicesOrIDs, size_t maxPoints) const override
void nn_radius_search(const mrpt::math::TPoint2Df &query, const float search_radius_sqr, std::vector<mrpt::math::TPoint2Df> &results, std::vector<float> &out_dists_sqr, std::vector<uint64_t> &resultIndicesOrIDs, size_t maxPoints) const override
template<size_t MAX_KNN>
void nn_multiple_search_impl(const mrpt::math::TPoint3Df &query, const size_t N, std::vector<mrpt::math::TPoint3Df> &results, std::vector<float> &out_dists_sqr, std::vector<uint64_t> &resultIndicesOrIDs) const

Public virtual methods implementation for NearestPlaneCapable

NearestPlaneResult nn_search_pt2pl(const mrpt::math::TPoint3Df &point, const float max_search_distance) const override

Public virtual methods implementation for CMetricMap

std::string asString() const override

Returns a short description of the map.

void getVisualizationInto(mrpt::opengl::CSetOfObjects &outObj) const override
bool isEmpty() const override

Returns true if the map is empty

void saveMetricMapRepresentationToFile(const std::string &filNamePrefix) const override

This virtual method saves the map to a file “filNamePrefix”+< some_file_extension >, as an image or in any other applicable way (Notice that other methods to save the map may be implemented in classes implementing this virtual interface).

const mrpt::maps::CSimplePointsMap *getAsSimplePointsMap() const override

Returns a cached point cloud view of the hash map. Not efficient at all. Only for MOLA->ROS2 bridge.

Public Types

using global_index3d_t = index3d_t<int32_t>

Public Members

TInsertionOptions insertionOptions
TLikelihoodOptions likelihoodOptions
TRenderOptions renderOptions
float voxel_size = 1.00f
mola::NDT::TInsertionOptions insertionOpts
mola::NDT::TLikelihoodOptions likelihoodOpts
mola::NDT::TRenderOptions renderOpts

Protected Functions

void internal_clear() override
struct point_vector_t

Public Members

std::array<float, MAX_POINTS_PER_VOXEL> xs
std::array<float, MAX_POINTS_PER_VOXEL> ys
std::array<float, MAX_POINTS_PER_VOXEL> zs
struct TInsertionOptions : public mrpt::config::CLoadableOptions

Options for insertObservation()

Public Functions

TInsertionOptions() = default
void loadFromConfigFile(const mrpt::config::CConfigFileBase &source, const std::string &section) override
void dumpToTextStream(std::ostream &out) const override
void writeToStream(mrpt::serialization::CArchive &out) const
void readFromStream(mrpt::serialization::CArchive &in)

Public Members

uint32_t max_points_per_voxel = 0

Maximum number of points per voxel. 0 means no limit further than the hard compile-time limit.

double remove_voxels_farther_than = .0

If !=0, remove the voxels farther (L1 distance) than this distance, in meters.

float min_distance_between_points = .0f

If !=0 skip the insertion of points that are closer than this distance to any other already in the voxel.

double max_eigen_ratio_for_planes = 0.01
struct TLikelihoodOptions : public mrpt::config::CLoadableOptions

Options used when evaluating “computeObservationLikelihood” in the derived classes.

See also

CObservation::computeObservationLikelihood

Public Functions

TLikelihoodOptions() = default
void loadFromConfigFile(const mrpt::config::CConfigFileBase &source, const std::string &section) override
void dumpToTextStream(std::ostream &out) const override
void writeToStream(mrpt::serialization::CArchive &out) const
void readFromStream(mrpt::serialization::CArchive &in)

Public Members

double sigma_dist = 0.5

Sigma (standard deviation, in meters) of the Gaussian observation model used to model the likelihood (default= 0.5 m)

double max_corr_distance = 1.0

Maximum distance in meters to consider for the numerator divided by “sigma_dist”, so that each point has a minimum (but very small) likelihood to avoid underflows (default=1.0 meters)

uint32_t decimation = 10

Speed up the likelihood computation by considering only one out of N rays (default=10)

struct TRenderOptions : public mrpt::config::CLoadableOptions

Rendering options, used in getAs3DObject()

Public Functions

void loadFromConfigFile(const mrpt::config::CConfigFileBase &source, const std::string &section) override
void dumpToTextStream(std::ostream &out) const override
void writeToStream(mrpt::serialization::CArchive &out) const

Binary dump to stream - used in derived classes’ serialization

void readFromStream(mrpt::serialization::CArchive &in)

Binary dump to stream - used in derived classes’ serialization

Public Members

bool points_visible = false
float point_size = 1.0f
mrpt::img::TColorf points_color = {.0f, .0f, 1.0f}
bool planes_visible = true
mrpt::img::TColorf planes_color = {1.0f, .0f, 1.0f}
bool normals_visible = true
mrpt::img::TColorf normals_color = {1.0f, 0.0f, 0.0f}
struct VoxelData

Public Functions

VoxelData() = default
inline auto points() const -> PointSpan
inline void insertPoint(const mrpt::math::TPoint3Df &p, const mrpt::math::TPoint3Df &sensorPose)
const std::optional<mp2p_icp::PointCloudEigen> &ndt() const

Gets cached NDT for this voxel, or computes it right now. If there are no points enough, nullopt is returned

inline bool has_ndt() const
inline size_t size() const
inline const auto &was_seen_from() const
struct PointSpan

Public Functions

inline PointSpan(const point_vector_t &points, uint32_t n)
inline size_t size() const
inline bool empty() const
inline mrpt::math::TPoint3Df operator[](int i) const