open3d.t.pipelines.registration.icp

open3d.t.pipelines.registration.icp(source, target, max_correspondence_distance, init_source_to_target=(with default value), estimation_method=TransformationEstimationPointToPoint, criteria=ICPConvergenceCriteria[relative_fitness_=1.000000e-06, relative_rmse=1.000000e-06, max_iteration_=30]., voxel_size=-1.0, save_loss_log=False)

Function for ICP registration

Parameters

• source (open3d.t.geometry.PointCloud) – The source point cloud.

• target (open3d.t.geometry.PointCloud) – The target point cloud.

• max_correspondence_distance (float) – Maximum correspondence points-pair distance.

• init_source_to_target (open3d.core.Tensor, optional) –

  Initial transformation estimation Default value:

  [[1.0 0.0 0.0 0.0], [0.0 1.0 0.0 0.0], [0.0 0.0 1.0 0.0], [0.0 0.0 0.0 1.0]] Tensor[shape={4, 4}, stride={4, 1}, Float64

  ()

• estimation_method (open3d.t.pipelines.registration.TransformationEstimation, optional, default=TransformationEstimationPointToPoint) – Estimation method. One of (TransformationEstimationPointToPoint, TransformationEstimationPointToPlane, TransformationEstimationForColoredICP, TransformationEstimationForGeneralizedICP)

• criteria (open3d.t.pipelines.registration.ICPConvergenceCriteria, optional, default=ICPConvergenceCriteria[relative_fitness_=1.000000e-06, relative_rmse=1.000000e-06, max_iteration_=30]) – Convergence criteria

• voxel_size (float, optional, default=-1.0) – 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 origianl scale of the point-cloud will be used.

• save_loss_log (bool, optional, default=False) – When True, it saves the iteration-wise values of fitness, inlier_rmse, transformaton, scale, iteration in loss_log_ in regsitration_result. Default: False.

Returns

open3d.t.pipelines.registration.RegistrationResult