# open3d.geometry.PointCloud¶

class open3d.geometry.PointCloud

PointCloud class. A point cloud consists of point coordinates, and optionally point colors and point normals.

class Type

Enum class for Geometry types.

HalfEdgeTriangleMesh = <Type.HalfEdgeTriangleMesh: 7>
Image = <Type.Image: 8>
LineSet = <Type.LineSet: 4>
PointCloud = <Type.PointCloud: 1>
RGBDImage = <Type.RGBDImage: 9>
TetraMesh = <Type.TetraMesh: 10>
TriangleMesh = <Type.TriangleMesh: 6>
Unspecified = <Type.Unspecified: 0>
VoxelGrid = <Type.VoxelGrid: 2>
property value
__init__(*args, **kwargs)

Overloaded function.

1. __init__(self: open3d.cpu.pybind.geometry.PointCloud) -> None

Default constructor

1. __init__(self: open3d.cpu.pybind.geometry.PointCloud, arg0: open3d.cpu.pybind.geometry.PointCloud) -> None

Copy constructor

1. __init__(self: open3d.cpu.pybind.geometry.PointCloud, points: open3d.cpu.pybind.utility.Vector3dVector) -> None

Create a PointCloud from points

clear(self)

Clear all elements in the geometry.

Returns

open3d.geometry.Geometry

cluster_dbscan(self, eps, min_points, print_progress=False)

Cluster PointCloud using the DBSCAN algorithm Ester et al., ‘A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases with Noise’, 1996. Returns a list of point labels, -1 indicates noise according to the algorithm.

Parameters
• eps (float) – Density parameter that is used to find neighbouring points.

• min_points (int) – Minimum number of points to form a cluster.

• print_progress (bool, optional, default=False) – If true the progress is visualized in the console.

Returns

open3d.utility.IntVector

compute_convex_hull(self: open3d.cpu.pybind.geometry.PointCloud, joggle_inputs: bool = False) → Tuple[open3d::geometry::TriangleMesh, List[int]]

Computes the convex hull of the point cloud.

Parameters

joggle_inputs (bool) – If True allows the algorithm to add random noise to the points to work around degenerate inputs. This adds the ‘QJ’ option to the qhull command.

Returns

The triangle mesh of the convex hull and the list of point indices that are part of the convex hull.

Return type

tuple(open3d.geometry.TriangleMesh, list)

compute_mahalanobis_distance(self)

Function to compute the Mahalanobis distance for points in a point cloud. See: https://en.wikipedia.org/wiki/Mahalanobis_distance.

Returns

open3d.utility.DoubleVector

compute_mean_and_covariance(self)

Function to compute the mean and covariance matrix of a point cloud.

Returns

Tuple[numpy.ndarray[numpy.float64[3, 1]], numpy.ndarray[numpy.float64[3, 3]]]

compute_nearest_neighbor_distance(self)

Function to compute the distance from a point to its nearest neighbor in the point cloud

Returns

open3d.utility.DoubleVector

compute_point_cloud_distance(self, target)

For each point in the source point cloud, compute the distance to the target point cloud.

Parameters

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

Returns

open3d.utility.DoubleVector

static create_from_depth_image(depth, intrinsic, extrinsic=(with default value), depth_scale=1000.0, depth_trunc=1000.0, stride=1, project_valid_depth_only=True)

Factory function to create a pointcloud from a depth image and a camera. Given depth value d at (u, v) image coordinate, the corresponding 3d point is:

• z = d / depth_scale

• x = (u - cx) * z / fx

• y = (v - cy) * z / fy

Parameters
• depth (open3d.geometry.Image) – The input depth image can be either a float image, or a uint16_t image.

• intrinsic (open3d.camera.PinholeCameraIntrinsic) – Intrinsic parameters of the camera.

• extrinsic (numpy.ndarray[numpy.float64[4, 4]], optional) – array([[1., 0., 0., 0.], [0., 1., 0., 0.], [0., 0., 1., 0.], [0., 0., 0., 1.]])

• depth_scale (float, optional, default=1000.0) – The depth is scaled by 1 / depth_scale.

• depth_trunc (float, optional, default=1000.0) – Truncated at depth_trunc distance.

• stride (int, optional, default=1) – Sampling factor to support coarse point cloud extraction.

• project_valid_depth_only (bool, optional, default=True) –

Returns

open3d.geometry.PointCloud

static create_from_rgbd_image(image, intrinsic, extrinsic=(with default value), project_valid_depth_only=True)

Factory function to create a pointcloud from an RGB-D image and a camera. Given depth value d at (u, v) image coordinate, the corresponding 3d point is:

• z = d / depth_scale

• x = (u - cx) * z / fx

• y = (v - cy) * z / fy

Parameters
• image (open3d.geometry.RGBDImage) – The input image.

• intrinsic (open3d.camera.PinholeCameraIntrinsic) – Intrinsic parameters of the camera.

• extrinsic (numpy.ndarray[numpy.float64[4, 4]], optional) – array([[1., 0., 0., 0.], [0., 1., 0., 0.], [0., 0., 1., 0.], [0., 0., 0., 1.]])

• project_valid_depth_only (bool, optional, default=True) –

Returns

open3d.geometry.PointCloud

crop(*args, **kwargs)

Overloaded function.

1. crop(self, bounding_box)

Function to crop input pointcloud into output pointcloud

Parameters

bounding_box (open3d.geometry.AxisAlignedBoundingBox) – AxisAlignedBoundingBox to crop points

Returns

open3d.geometry.PointCloud

1. crop(self, bounding_box)

Function to crop input pointcloud into output pointcloud

Parameters

bounding_box (open3d.geometry.OrientedBoundingBox) – AxisAlignedBoundingBox to crop points

Returns

open3d.geometry.PointCloud

detect_planar_patches(self, normal_variance_threshold_deg=60, coplanarity_deg=75, outlier_ratio=0.75, min_plane_edge_length=0.0, min_num_points=0, search_param=KDTreeSearchParamKNN with knn = 30)

Detects planar patches in the point cloud using a robust statistics-based approach.

Returns

A list of detected planar patches, represented as OrientedBoundingBox objects, with the third column (z) of R indicating the planar patch normal vector. The extent in the z direction is non-zero so that the OrientedBoundingBox contains the points that contribute to the plane detection.

Parameters
• normal_variance_threshold_deg (float, optional, default=60) –

• coplanarity_deg (float, optional, default=75) –

• outlier_ratio (float, optional, default=0.75) – Maximum allowable ratio of outliers associated to a plane.

• min_plane_edge_length (float, optional, default=0.0) – Minimum edge length of plane’s long edge before being rejected.

• min_num_points (int, optional, default=0) – Minimum number of points allowable for fitting planes.

• search_param (open3d.geometry.KDTreeSearchParam, optional, default=KDTreeSearchParamKNN with knn = 30) – The KDTree search parameters for neighborhood search.

Returns

List[open3d.geometry.OrientedBoundingBox]

dimension(self)

Returns whether the geometry is 2D or 3D.

Returns

int

estimate_covariances(self, search_param=KDTreeSearchParamKNN with knn = 30)

Function to compute the covariance matrix for each point in the point cloud

Parameters

search_param (open3d.geometry.KDTreeSearchParam, optional, default=KDTreeSearchParamKNN with knn = 30) – The KDTree search parameters for neighborhood search.

Returns

None

estimate_normals(self, search_param=KDTreeSearchParamKNN with knn = 30, fast_normal_computation=True)

Function to compute the normals of a point cloud. Normals are oriented with respect to the input point cloud if normals exist

Parameters
• search_param (open3d.geometry.KDTreeSearchParam, optional, default=KDTreeSearchParamKNN with knn = 30) – The KDTree search parameters for neighborhood search.

• fast_normal_computation (bool, optional, default=True) – If true, the normal estimation uses a non-iterative method to extract the eigenvector from the covariance matrix. This is faster, but is not as numerical stable.

Returns

None

static estimate_point_covariances(input, search_param=KDTreeSearchParamKNN with knn = 30)

Static function to compute the covariance matrix for each point in the given point cloud, doesn’t change the input

Parameters
Returns

open3d.utility.Matrix3dVector

farthest_point_down_sample(self: open3d.cpu.pybind.geometry.PointCloud, num_samples: int) → open3d.cpu.pybind.geometry.PointCloud

Downsamples input pointcloud into output pointcloud with a set of points has farthest distance. The sample is performed by selecting the farthest point from previous selected points iteratively.

get_axis_aligned_bounding_box(self)

Returns an axis-aligned bounding box of the geometry.

Returns

open3d.geometry.AxisAlignedBoundingBox

get_center(self)

Returns the center of the geometry coordinates.

Returns

numpy.ndarray[numpy.float64[3, 1]]

get_geometry_type(self)

Returns one of registered geometry types.

Returns

open3d.geometry.Geometry.GeometryType

get_max_bound(self)

Returns max bounds for geometry coordinates.

Returns

numpy.ndarray[numpy.float64[3, 1]]

get_min_bound(self)

Returns min bounds for geometry coordinates.

Returns

numpy.ndarray[numpy.float64[3, 1]]

get_minimal_oriented_bounding_box(self: open3d.cpu.pybind.geometry.Geometry3D, robust: bool = False) → open3d::geometry::OrientedBoundingBox

Returns the minimal oriented bounding box for the geometry.

Creates the oriented bounding box with the smallest volume. The algorithm makes use of the fact that at least one edge of the convex hull must be collinear with an edge of the minimum bounding box: for each triangle in the convex hull, calculate the minimal axis aligned box in the frame of that triangle. at the end, return the box with the smallest volume

Parameters

robust (bool) – If set to true uses a more robust method which works in degenerate cases but introduces noise to the points coordinates.

Returns

The oriented bounding box. The bounding box is oriented such that its volume is minimized.

Return type

open3d.geometry.OrientedBoundingBox

get_oriented_bounding_box(self: open3d.cpu.pybind.geometry.Geometry3D, robust: bool = False) → open3d::geometry::OrientedBoundingBox

Returns the oriented bounding box for the geometry.

Computes the oriented bounding box based on the PCA of the convex hull. The returned bounding box is an approximation to the minimal bounding box.

Parameters

robust (bool) – If set to true uses a more robust method which works in degenerate cases but introduces noise to the points coordinates.

Returns

The oriented bounding box. The bounding box is oriented such that the axes are ordered with respect to the principal components.

Return type

open3d.geometry.OrientedBoundingBox

static get_rotation_matrix_from_axis_angle(rotation: numpy.ndarray[numpy.float64[3, 1]]) → numpy.ndarray[numpy.float64[3, 3]]
static get_rotation_matrix_from_quaternion(rotation: numpy.ndarray[numpy.float64[4, 1]]) → numpy.ndarray[numpy.float64[3, 3]]
static get_rotation_matrix_from_xyz(rotation: numpy.ndarray[numpy.float64[3, 1]]) → numpy.ndarray[numpy.float64[3, 3]]
static get_rotation_matrix_from_xzy(rotation: numpy.ndarray[numpy.float64[3, 1]]) → numpy.ndarray[numpy.float64[3, 3]]
static get_rotation_matrix_from_yxz(rotation: numpy.ndarray[numpy.float64[3, 1]]) → numpy.ndarray[numpy.float64[3, 3]]
static get_rotation_matrix_from_yzx(rotation: numpy.ndarray[numpy.float64[3, 1]]) → numpy.ndarray[numpy.float64[3, 3]]
static get_rotation_matrix_from_zxy(rotation: numpy.ndarray[numpy.float64[3, 1]]) → numpy.ndarray[numpy.float64[3, 3]]
static get_rotation_matrix_from_zyx(rotation: numpy.ndarray[numpy.float64[3, 1]]) → numpy.ndarray[numpy.float64[3, 3]]
has_colors(self)

Returns True if the point cloud contains point colors.

Returns

bool

has_covariances(self: open3d.cpu.pybind.geometry.PointCloud) → bool

Returns True if the point cloud contains covariances.

has_normals(self)

Returns True if the point cloud contains point normals.

Returns

bool

has_points(self)

Returns True if the point cloud contains points.

Returns

bool

hidden_point_removal(self, camera_location, radius)

Removes hidden points from a point cloud and returns a mesh of the remaining points. Based on Katz et al. ‘Direct Visibility of Point Sets’, 2007. Additional information about the choice of radius for noisy point clouds can be found in Mehra et. al. ‘Visibility of Noisy Point Cloud Data’, 2010.

Parameters
• camera_location (numpy.ndarray[numpy.float64[3, 1]]) – All points not visible from that location will be removed

• radius (float) – The radius of the sperical projection

Returns

Tuple[open3d.geometry.TriangleMesh, List[int]]

is_empty(self)

Returns True iff the geometry is empty.

Returns

bool

normalize_normals(self)

Normalize point normals to length 1.

Returns

open3d.geometry.PointCloud

orient_normals_consistent_tangent_plane(self, k)

Function to orient the normals with respect to consistent tangent planes

Parameters

k (int) – Number of k nearest neighbors used in constructing the Riemannian graph used to propagate normal orientation.

Returns

None

orient_normals_to_align_with_direction(self, orientation_reference=array([0.0, 0.0, 1.0]))

Function to orient the normals of a point cloud

Parameters

orientation_reference (numpy.ndarray[numpy.float64[3, 1]], optional, default=array([0., 0., 1.])) – Normals are oriented with respect to orientation_reference.

Returns

None

orient_normals_towards_camera_location(self, camera_location=array([0.0, 0.0, 0.0]))

Function to orient the normals of a point cloud

Parameters

camera_location (numpy.ndarray[numpy.float64[3, 1]], optional, default=array([0., 0., 0.])) – Normals are oriented with towards the camera_location.

Returns

None

paint_uniform_color(self, color)

Assigns each point in the PointCloud the same color.

Parameters

color (numpy.ndarray[numpy.float64[3, 1]]) – RGB color for the PointCloud.

Returns

open3d.geometry.PointCloud

random_down_sample(self, sampling_ratio)

Function to downsample input pointcloud into output pointcloud randomly. The sample is generated by randomly sampling the indexes from the point cloud.

Parameters

sampling_ratio (float) – Sampling ratio, the ratio of number of selected points to total number of points[0-1]

Returns

open3d.geometry.PointCloud

remove_duplicated_points(self)

Removes duplicated points, i.e., points that have identical coordinates. It also removes the corresponding attributes associated with the non-finite point such as normals, covariances and color entries. It doesn’t re-computes these attributes after removing duplicated points.

Returns

open3d.geometry.PointCloud

remove_non_finite_points(self, remove_nan=True, remove_infinite=True)

Removes all points from the point cloud that have a nan entry, or infinite entries. It also removes the corresponding attributes associated with the non-finite point such as normals, covariances and color entries. It doesn’t re-computes these attributes after removing non-finite points.

Parameters
• remove_nan (bool, optional, default=True) – Remove NaN values from the PointCloud

• remove_infinite (bool, optional, default=True) – Remove infinite values from the PointCloud

Returns

open3d.geometry.PointCloud

remove_radius_outlier(self, nb_points, radius, print_progress=False)

Removes points that have neighbors less than nb_points in a sphere of a given radius

Parameters
• nb_points (int) – Number of points within the radius.

• radius (float) – Radius of the sphere.

• print_progress (bool, optional, default=False) – Set to True to print progress bar.

Returns

Tuple[open3d.geometry.PointCloud, List[int]]

remove_statistical_outlier(self, nb_neighbors, std_ratio, print_progress=False)

Removes points that are further away from their neighbors in average.

Parameters
• nb_neighbors (int) – Number of neighbors around the target point.

• std_ratio (float) – Standard deviation ratio.

• print_progress (bool, optional, default=False) – Set to True to print progress bar.

Returns

Tuple[open3d.geometry.PointCloud, List[int]]

rotate(*args, **kwargs)

Overloaded function.

1. rotate(self, R)

Apply rotation to the geometry coordinates and normals.

Parameters

R (numpy.ndarray[numpy.float64[3, 3]]) – The rotation matrix

Returns

open3d.geometry.Geometry3D

1. rotate(self, R, center)

Apply rotation to the geometry coordinates and normals.

Parameters
• R (numpy.ndarray[numpy.float64[3, 3]]) – The rotation matrix

• center (numpy.ndarray[numpy.float64[3, 1]]) – Rotation center used for transformation.

Returns

open3d.geometry.Geometry3D

scale(*args, **kwargs)

Overloaded function.

1. scale(self, scale, center)

Apply scaling to the geometry coordinates.

Parameters
• scale (float) – The scale parameter that is multiplied to the points/vertices of the geometry.

• center (numpy.ndarray[numpy.float64[3, 1]]) – Scale center used for transformation.

Returns

open3d.geometry.Geometry3D

1. scale(self, scale, center)

Apply scaling to the geometry coordinates.

Parameters
• scale (float) – The scale parameter that is multiplied to the points/vertices of the geometry.

• center (numpy.ndarray[numpy.float64[3, 1]]) – Scale center used for transformation.

Returns

open3d.geometry.Geometry3D

segment_plane(self, distance_threshold, ransac_n, num_iterations, probability=0.99999999)

Segments a plane in the point cloud using the RANSAC algorithm.

Parameters
• distance_threshold (float) – Max distance a point can be from the plane model, and still be considered an inlier.

• ransac_n (int) – Number of initial points to be considered inliers in each iteration.

• num_iterations (int) – Number of iterations.

• probability (float, optional, default=0.99999999) – Expected probability of finding the optimal plane.

Returns

Tuple[numpy.ndarray[numpy.float64[4, 1]], List[int]]

select_by_index(self, indices, invert=False)

Function to select points from input pointcloud into output pointcloud.

Parameters
• indices (List[int]) – Indices of points to be selected.

• invert (bool, optional, default=False) – Set to True to invert the selection of indices.

Returns

open3d.geometry.PointCloud

transform(self, arg0)

Apply transformation (4x4 matrix) to the geometry coordinates.

Parameters

arg0 (numpy.ndarray[numpy.float64[4, 4]]) –

Returns

open3d.geometry.Geometry3D

translate(self, translation, relative=True)

Apply translation to the geometry coordinates.

Parameters
• translation (numpy.ndarray[numpy.float64[3, 1]]) – A 3D vector to transform the geometry

• relative (bool, optional, default=True) – If true, the translation vector is directly added to the geometry coordinates. Otherwise, the center is moved to the translation vector.

Returns

open3d.geometry.Geometry3D

uniform_down_sample(self, every_k_points)

Function to downsample input pointcloud into output pointcloud uniformly. The sample is performed in the order of the points with the 0-th point always chosen, not at random.

Parameters

every_k_points (int) – Sample rate, the selected point indices are [0, k, 2k, …]

Returns

open3d.geometry.PointCloud

voxel_down_sample(self, voxel_size)

Function to downsample input pointcloud into output pointcloud with a voxel. Normals and colors are averaged if they exist.

Parameters

voxel_size (float) – Voxel size to downsample into.

Returns

open3d.geometry.PointCloud

voxel_down_sample_and_trace(self, voxel_size, min_bound, max_bound, approximate_class=False)

Function to downsample using PointCloud.VoxelDownSample. Also records point cloud index before downsampling

Parameters
• voxel_size (float) – Voxel size to downsample into.

• min_bound (numpy.ndarray[numpy.float64[3, 1]]) – Minimum coordinate of voxel boundaries

• max_bound (numpy.ndarray[numpy.float64[3, 1]]) – Maximum coordinate of voxel boundaries

• approximate_class (bool, optional, default=False) –

Returns

Tuple[open3d.geometry.PointCloud, numpy.ndarray[numpy.int32[m, n]], List[open3d.utility.IntVector]]

HalfEdgeTriangleMesh = <Type.HalfEdgeTriangleMesh: 7>
Image = <Type.Image: 8>
LineSet = <Type.LineSet: 4>
PointCloud = <Type.PointCloud: 1>
RGBDImage = <Type.RGBDImage: 9>
TetraMesh = <Type.TetraMesh: 10>
TriangleMesh = <Type.TriangleMesh: 6>
Unspecified = <Type.Unspecified: 0>
VoxelGrid = <Type.VoxelGrid: 2>
property colors

RGB colors of points.

Type

float64 array of shape (num_points, 3), range [0, 1] , use numpy.asarray() to access data

property covariances

Points covariances.

Type

float64 array of shape (num_points, 3, 3), use numpy.asarray() to access data

property normals

Points normals.

Type

float64 array of shape (num_points, 3), use numpy.asarray() to access data

property points

Points coordinates.

Type

float64 array of shape (num_points, 3), use numpy.asarray() to access data