open3d.geometry.VoxelGrid#

class open3d.geometry.VoxelGrid#

VoxelGrid is a collection of voxels which are aligned in grid.

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.VoxelGrid) -> None

Default constructor

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

Copy constructor

add_voxel(self, voxel)#

Add a new voxel into the VoxelGrid.

Parameters:

voxel (open3d.geometry.Voxel) –

Returns:

None

carve_depth_map(self, depth_map, camera_params, keep_voxels_outside_image=False)#

Remove all voxels from the VoxelGrid where none of the boundary points of the voxel projects to depth value that is smaller, or equal than the projected depth of the boundary point. If keep_voxels_outside_image is true then voxels are only carved if all boundary points project to a valid image location.

Parameters:
Returns:

open3d.geometry.VoxelGrid

carve_silhouette(self, silhouette_mask, camera_params, keep_voxels_outside_image=False)#

Remove all voxels from the VoxelGrid where none of the boundary points of the voxel projects to a valid mask pixel (pixel value > 0). If keep_voxels_outside_image is true then voxels are only carved if all boundary points project to a valid image location.

Parameters:
Returns:

open3d.geometry.VoxelGrid

check_if_included(self, queries)#

Element-wise check if a query in the list is included in the VoxelGrid. Queries are double precision and are mapped to the closest voxel.

Parameters:

queries (open3d.utility.Vector3dVector) –

Returns:

List[bool]

clear(self)#

Clear all elements in the geometry.

Returns:

open3d.geometry.Geometry

static create_dense(origin, color, voxel_size, width, height, depth)#

Creates a voxel grid where every voxel is set (hence dense). This is a useful starting point for voxel carving

Parameters:

  • origin (numpy.ndarray[numpy.float64[3, 1]]) – Coordinate center of the VoxelGrid

  • color (numpy.ndarray[numpy.float64[3, 1]]) – Voxel color for all voxels if the VoxelGrid.

  • voxel_size (float) – Voxel size of of the VoxelGrid construction.

  • width (float) – Spatial width extend of the VoxelGrid.

  • height (float) – Spatial height extend of the VoxelGrid.

  • depth (float) – Spatial depth extend of the VoxelGrid.

Returns:

open3d.geometry.VoxelGrid

create_from_octree(self, octree)#

Convert from Octree.

Parameters:

octree (open3d.geometry.Octree) – geometry.Octree: The source octree.

Returns:

None

static create_from_point_cloud(input, voxel_size)#

Creates a VoxelGrid from a given PointCloud. The color value of a given voxel is the average color value of the points that fall into it (if the PointCloud has colors). The bounds of the created VoxelGrid are computed from the PointCloud.

Parameters:
Returns:

open3d.geometry.VoxelGrid

static create_from_point_cloud_within_bounds(input, voxel_size, min_bound, max_bound)#

Creates a VoxelGrid from a given PointCloud. The color value of a given voxel is the average color value of the points that fall into it (if the PointCloud has colors). The bounds of the created VoxelGrid are defined by the given parameters.

Parameters:

  • input (open3d.geometry.PointCloud) – The input PointCloud

  • voxel_size (float) – Voxel size of of the VoxelGrid construction.

  • min_bound (numpy.ndarray[numpy.float64[3, 1]]) – Minimum boundary point for the VoxelGrid to create.

  • max_bound (numpy.ndarray[numpy.float64[3, 1]]) – Maximum boundary point for the VoxelGrid to create.

Returns:

open3d.geometry.VoxelGrid

static create_from_triangle_mesh(input, voxel_size)#

Creates a VoxelGrid from a given TriangleMesh. No color information is converted. The bounds of the created VoxelGrid are computed from the TriangleMesh.

Parameters:
Returns:

open3d.geometry.VoxelGrid

static create_from_triangle_mesh_within_bounds(input, voxel_size, min_bound, max_bound)#

Creates a VoxelGrid from a given TriangleMesh. No color information is converted. The bounds of the created VoxelGrid are defined by the given parameters

Parameters:

  • input (open3d.geometry.TriangleMesh) – The input TriangleMesh

  • voxel_size (float) – Voxel size of of the VoxelGrid construction.

  • min_bound (numpy.ndarray[numpy.float64[3, 1]]) – Minimum boundary point for the VoxelGrid to create.

  • max_bound (numpy.ndarray[numpy.float64[3, 1]]) – Maximum boundary point for the VoxelGrid to create.

Returns:

open3d.geometry.VoxelGrid

dimension(self)#

Returns whether the geometry is 2D or 3D.

Returns:

int

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]]#
get_voxel(self, point)#

Returns voxel index given query point.

Parameters:

point (numpy.ndarray[numpy.float64[3, 1]]) – The query point.

Returns:

numpy.ndarray[numpy.int32[3, 1]]

get_voxel_bounding_points(self, index)#

Returns the 8 bounding points of a voxel given its grid index.

Parameters:

index (numpy.ndarray[numpy.int32[3, 1]]) – The grid index of the query voxel.

Returns:

open3d.utility.Vector3dVector

get_voxel_center_coordinate(self, idx)#

Returns the center coordinate of a voxel given its grid index.

Parameters:

idx (numpy.ndarray[numpy.int32[3, 1]]) – The grid index of the query voxel.

Returns:

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

get_voxels(self: open3d.cpu.pybind.geometry.VoxelGrid) → List[open3d.cpu.pybind.geometry.Voxel]#

Returns List of Voxel: Voxels contained in voxel grid. Changes to the voxels returned from this methodare not reflected in the voxel grid.

has_colors(self)#

Returns True if the voxel grid contains voxel colors.

Returns:

bool

has_voxels(self)#

Returns True if the voxel grid contains voxels.

Returns:

bool

is_empty(self)#

Returns True iff the geometry is empty.

Returns:

bool

remove_voxel(self, idx)#

Remove a voxel given index.

Parameters:

idx (numpy.ndarray[numpy.int32[3, 1]]) – The grid index of the target voxel.

Returns:

None

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

to_octree(self, max_depth)#

Convert to Octree.

Parameters:

max_depth (int) – int: Maximum depth of the octree.

Returns:

open3d.geometry.Octree

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

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 origin#

Coordinate of the origin point.

Type:

float64 vector of length 3

property voxel_size#

float64 Size of the voxel.