open3d::t::pipelines::registration Namespace Reference

Data Structures
class	ICPConvergenceCriteria
	Class that defines the convergence criteria of ICP. More...
class	RegistrationResult
class	RobustKernel
class	TransformationEstimation
class	TransformationEstimationPointToPoint
class	TransformationEstimationPointToPlane
class	TransformationEstimationForColoredICP
class	TransformationEstimationForDopplerICP

Enumerations
enum class	RobustKernelMethod {   L2Loss = 0 , L1Loss = 1 , HuberLoss = 2 , CauchyLoss = 3 ,   GMLoss = 4 , TukeyLoss = 5 , GeneralizedLoss = 6 }
enum class	TransformationEstimationType {   Unspecified = 0 , PointToPoint = 1 , PointToPlane = 2 , ColoredICP = 3 ,   DopplerICP = 4 }

Functions
core::Tensor	ComputeFPFHFeature (const geometry::PointCloud &input, const utility::optional< int > max_nn, const utility::optional< double > radius)
core::Tensor	CorrespondencesFromFeatures (const core::Tensor &source_features, const core::Tensor &target_features, bool mutual_filter=false, float mutual_consistency_ratio=0.1)
	Function to find correspondences via 1-nearest neighbor feature matching. Target is used to construct a nearest neighbor search object, in order to query source. More...
RegistrationResult	EvaluateRegistration (const geometry::PointCloud &source, const geometry::PointCloud &target, double max_correspondence_distance, const core::Tensor &transformation=core::Tensor::Eye(4, core::Float64, core::Device("CPU:0")))
	Function for evaluating registration between point clouds. More...
RegistrationResult	ICP (const geometry::PointCloud &source, const geometry::PointCloud &target, const double max_correspondence_distance, const core::Tensor &init_source_to_target=core::Tensor::Eye(4, core::Float64, core::Device("CPU:0")), const TransformationEstimation &estimation=TransformationEstimationPointToPoint(), const ICPConvergenceCriteria &criteria=ICPConvergenceCriteria(), const double voxel_size=-1.0, const std::function< void(const std::unordered_map< std::string, core::Tensor > &)> &callback_after_iteration=nullptr)
	Functions for ICP registration. More...
RegistrationResult	MultiScaleICP (const geometry::PointCloud &source, const geometry::PointCloud &target, const std::vector< double > &voxel_sizes, const std::vector< ICPConvergenceCriteria > &criteria_list, const std::vector< double > &max_correspondence_distances, const core::Tensor &init_source_to_target=core::Tensor::Eye(4, core::Float64, core::Device("CPU:0")), const TransformationEstimation &estimation=TransformationEstimationPointToPoint(), const std::function< void(const std::unordered_map< std::string, core::Tensor > &)> &callback_after_iteration=nullptr)
	Functions for Multi-Scale ICP registration. It will run ICP on different voxel level, from coarse to dense. The vector of ICPConvergenceCriteria(relative fitness, relative rmse, max_iterations) contains the stopping condition for each voxel level. The length of voxel_sizes vector, criteria vector, max_correspondence_distances vector must be same, and voxel_sizes must contain positive values in strictly decreasing order [Lower the voxel size, higher is the resolution]. Only the last value of the voxel_sizes vector can be {-1}, as it allows to run on the original scale without downsampling. More...
core::Tensor	GetInformationMatrix (const geometry::PointCloud &source, const geometry::PointCloud &target, const double max_correspondence_distance, const core::Tensor &transformation)
	Computes Information Matrix, from the transfromation between source and target pointcloud. It returns the Information Matrix of shape {6, 6}, of dtype Float64 on device CPU:0. More...

Enumeration Type Documentation

RobustKernelMethod

enum open3d::t::pipelines::registration::RobustKernelMethod

strong

Enumerator
L2Loss
L1Loss
HuberLoss
CauchyLoss
GMLoss
TukeyLoss
GeneralizedLoss

TransformationEstimationType

enum open3d::t::pipelines::registration::TransformationEstimationType

strong

Enumerator
Unspecified
PointToPoint
PointToPlane
ColoredICP
DopplerICP

Function Documentation

ComputeFPFHFeature()

core::Tensor open3d::t::pipelines::registration::ComputeFPFHFeature	(	const geometry::PointCloud &	input,
		const utility::optional< int >	max_nn = 100,
		const utility::optional< double >	radius = utility::nullopt
	)

Function to compute FPFH feature for a point cloud. It uses KNN search (Not recommended to use on GPU) if only max_nn parameter is provided, Radius search (Not recommended to use on GPU) if only radius parameter is provided, and Hybrid search (Recommended) if both are provided.

Parameters

input	The input point cloud with data type float32 or float64.
max_nn	[optional] Neighbor search max neighbors parameter. [Default = 100].
radius	[optional] Neighbor search radius parameter. [Recommended ~5x voxel size].

Returns

A Tensor of FPFH feature of the input point cloud with shape {N, 33}, data type and device same as input.

CorrespondencesFromFeatures()

core::Tensor open3d::t::pipelines::registration::CorrespondencesFromFeatures	(	const core::Tensor &	source_features,
		const core::Tensor &	target_features,
		bool	mutual_filter = false,
		float	mutual_consistency_ratio = 0.1
	)

Function to find correspondences via 1-nearest neighbor feature matching. Target is used to construct a nearest neighbor search object, in order to query source.

Parameters

source_feats	(N, D) tensor
target_feats	(M, D) tensor
mutual_filter	Boolean flag, only return correspondences (i, j) s.t. source_features[i] and target_features[j] are mutually the nearest neighbor.
mutual_consistency_ratio	Float threshold to decide whether the number of correspondences is sufficient. Only used when mutual_filter is set to True.

Returns

(K, 2, Int64) tensor. When mutual_filter is disabled: the first column is arange(0, N) of source, and the second column is the corresponding index of target. When mutual_filter is enabled, return the filtering subset of the aforementioned correspondence set where source[i] and target[j] are mutually the nearest neighbor. If the subset size is smaller than mutual_consistency_ratio * N, return the unfiltered set.

EvaluateRegistration()

RegistrationResult open3d::t::pipelines::registration::EvaluateRegistration	(	const geometry::PointCloud &	source,
		const geometry::PointCloud &	target,
		double	max_correspondence_distance,
		const core::Tensor &	transformation = core::Tensor::Eye(4, core::Float64, core::Device("CPU:0"))
	)

Function for evaluating registration between point clouds.

Parameters

source	The source point cloud. (Float32 or Float64 type).
target	The target point cloud. (Float32 or Float64 type).
max_correspondence_distance	Maximum correspondence points-pair distance.
transformation	The 4x4 transformation matrix to transform source to target of dtype Float64 on CPU device.

GetInformationMatrix()

core::Tensor open3d::t::pipelines::registration::GetInformationMatrix	(	const geometry::PointCloud &	source,
		const geometry::PointCloud &	target,
		const double	max_correspondence_distance,
		const core::Tensor &	transformation
	)

Computes Information Matrix, from the transfromation between source and target pointcloud. It returns the Information Matrix of shape {6, 6}, of dtype Float64 on device CPU:0.

Parameters

source	The source point cloud. (Float32 or Float64 type).
target	The target point cloud. (Float32 or Float64 type).
max_correspondence_distance	Maximum correspondence points-pair distance.
transformation	The 4x4 transformation matrix to transform source to target.

ICP()

RegistrationResult open3d::t::pipelines::registration::ICP	(	const geometry::PointCloud &	source,
		const geometry::PointCloud &	target,
		const double	max_correspondence_distance,
		const core::Tensor &	init_source_to_target = core::Tensor::Eye(4, core::Float64, core::Device("CPU:0")),
		const TransformationEstimation &	estimation = TransformationEstimationPointToPoint(),
		const ICPConvergenceCriteria &	criteria = ICPConvergenceCriteria(),
		const double	voxel_size = -1.0,
		const std::function< void(const std::unordered_map< std::string, core::Tensor > &)> &	callback_after_iteration = nullptr
	)

Functions for ICP registration.

Parameters

source	The source point cloud. (Float32 or Float64 type).
target	The target point cloud. (Float32 or Float64 type).
max_correspondence_distance	Maximum correspondence points-pair distance.
init_source_to_target	Initial transformation estimation of type Float64 on CPU.
estimation	Estimation method.
criteria	Convergence criteria.
voxel_size	The input pointclouds will be down-sampled to this voxel_size scale. If voxel_size < 0, original scale will be used. However it is highly recommended to down-sample the point-cloud for performance. By default original scale of the point-cloud will be used.
callback_after_iteration	Optional lambda function, saves string to tensor map of attributes such as "iteration_index", "scale_index", "scale_iteration_index", "inlier_rmse", "fitness", "transformation", on CPU device, updated after each iteration.

MultiScaleICP()

RegistrationResult open3d::t::pipelines::registration::MultiScaleICP	(	const geometry::PointCloud &	source,
		const geometry::PointCloud &	target,
		const std::vector< double > &	voxel_sizes,
		const std::vector< ICPConvergenceCriteria > &	criteria_list,
		const std::vector< double > &	max_correspondence_distances,
		const core::Tensor &	init_source_to_target = core::Tensor::Eye(4, core::Float64, core::Device("CPU:0")),
		const TransformationEstimation &	estimation = TransformationEstimationPointToPoint(),
		const std::function< void(const std::unordered_map< std::string, core::Tensor > &)> &	callback_after_iteration = nullptr
	)

Functions for Multi-Scale ICP registration. It will run ICP on different voxel level, from coarse to dense. The vector of ICPConvergenceCriteria(relative fitness, relative rmse, max_iterations) contains the stopping condition for each voxel level. The length of voxel_sizes vector, criteria vector, max_correspondence_distances vector must be same, and voxel_sizes must contain positive values in strictly decreasing order [Lower the voxel size, higher is the resolution]. Only the last value of the voxel_sizes vector can be {-1}, as it allows to run on the original scale without downsampling.

Parameters

source	The source point cloud. (Float32 or Float64 type).
target	The target point cloud. (Float32 or Float64 type).
voxel_sizes	VectorDouble of voxel scales of type double.
criteria_list	Vector of ICPConvergenceCriteria objects for each scale.
max_correspondence_distances	VectorDouble of maximum correspondence points-pair distances of type double, for each iteration. Must be of same length as voxel_sizes and criterias.
init_source_to_target	Initial transformation estimation of type Float64 on CPU.
estimation	Estimation method.
callback_after_iteration	Optional lambda function, saves string to tensor map of attributes such as "iteration_index", "scale_index", "scale_iteration_index", "inlier_rmse", "fitness", "transformation", on CPU device, updated after each iteration.