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 1 0 0], [0 0 1 0], [0 0 0 1]] 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