open3d.pipelines.registration#

Classes

CauchyLoss

The loss \(\rho(r)\) for a given residual r is:

CorrespondenceChecker

Base class that checks if two (small) point clouds can be aligned.

CorrespondenceCheckerBasedOnDistance

Class to check if aligned point clouds are close (less than specified threshold).

CorrespondenceCheckerBasedOnEdgeLength

Check if two point clouds build the polygons with similar edge lengths.

CorrespondenceCheckerBasedOnNormal

Class to check if two aligned point clouds have similar normals.

FastGlobalRegistrationOption

Options for FastGlobalRegistration.

Feature

Class to store featrues for registration.

GMLoss

The loss \(\rho(r)\) for a given residual r is:

GlobalOptimizationConvergenceCriteria

Convergence criteria of GlobalOptimization.

GlobalOptimizationGaussNewton

Global optimization with Gauss-Newton algorithm.

GlobalOptimizationLevenbergMarquardt

Global optimization with Levenberg-Marquardt algorithm.

GlobalOptimizationMethod

Base class for global optimization method.

GlobalOptimizationOption

Option for GlobalOptimization.

HuberLoss

The loss \(\rho(r)\) for a given residual r is:

ICPConvergenceCriteria

Class that defines the convergence criteria of ICP.

L1Loss

The loss \(\rho(r)\) for a given residual r is given by:

L2Loss

The loss \(\rho(r)\) for a given residual r is given by:

PoseGraph

Data structure defining the pose graph.

PoseGraphEdge

Edge of PoseGraph.

PoseGraphEdgeVector

Vector of PoseGraphEdge

PoseGraphNode

Node of PoseGraph.

PoseGraphNodeVector

Vector of PoseGraphNode

RANSACConvergenceCriteria

Class that defines the convergence criteria of RANSAC.

RegistrationResult

Class that contains the registration results.

RobustKernel

Base class that models a robust kernel for outlier rejection.

TransformationEstimation

Base class that estimates a transformation between two point clouds.

TransformationEstimationForColoredICP

Class to estimate a transformation between two point clouds using color information

TransformationEstimationForGeneralizedICP

Class to estimate a transformation for Generalized ICP.

TransformationEstimationPointToPlane

Class to estimate a transformation for point to plane distance.

TransformationEstimationPointToPoint

Class to estimate a transformation for point to point distance.

TukeyLoss

The loss \(\rho(r)\) for a given residual r is:

Functions

compute_fpfh_feature(input, search_param)

Function to compute FPFH feature for a point cloud

correspondences_from_features(...[, ...])

Function to find nearest neighbor correspondences from features

evaluate_registration(source, target, ...[, ...])

Function for evaluating registration between point clouds

get_information_matrix_from_point_clouds(...)

Function for computing information matrix from transformation matrix

global_optimization(pose_graph, method, ...)

Function to optimize PoseGraph

registration_colored_icp(source, target, ...)

Function for Colored ICP registration

registration_fgr_based_on_correspondence(...)

Function for fast global registration based on a set of correspondences

registration_fgr_based_on_feature_matching(...)

Function for fast global registration based on feature matching

registration_generalized_icp(source, target, ...)

Function for Generalized ICP registration

registration_icp(source, target, ...[, ...])

Function for ICP registration

registration_ransac_based_on_correspondence(...)

Function for global RANSAC registration based on a set of correspondences

registration_ransac_based_on_feature_matching(...)

Function for global RANSAC registration based on feature matching