Utilities - GRID Documentation

General

get_public_fields

Get all public fields of an object (not starting with _).

Object

The object to retrieve fields from.

List[str]

list_to_2d_float_array

Convert a list of floats to a 2D array.

List[float]

List of float values.

int

The width of the array.

int

The height of the array.

np.ndarray

A NumPy array with dtype=np.float32 and shape (height, width).

string_to_float_array

Convert a string to a float array.

str | bytes

The string or bytes to convert.

np.ndarray

string_to_uint8_array

Convert a string to a uint8 array.

str | bytes

The string or bytes to convert.

np.ndarray

A NumPy array with dtype=uint8.

to_dict

Convert an object to a dictionary of its public fields.

Object

The object to convert.

Dict[str, Any]

to_str

Convert an object to a string of its public fields.

Object

The object to convert.

str

vector3d2list

Turn an AirGen Vector3r into a list of three floats.

Vector3r

The 3D vector to convert.

List[float]

wait_key

Wait for a key press on the console and return it.

str, optional

message to display while waiting.

write_file

Write binary data to a file. Used for writing compressed PNG images.

str

The name of the file.

bytes

The binary data to write.

Sensor

imagetype2request

Helper function to create an uncompressed AirGen ImageRequest for a given image type.

Literal["front_center", "bottom_center"]

The name of the camera.

ImageType

The type of image to capture.

ImageRequest

responses2images

Convert AirGen image responses to NumPy arrays.

List[ImageResponse]

The image responses to convert.

Optional[Any]

The data type for the resulting NumPy arrays.

Dict[str, np.ndarray]

A dictionary with keys corresponding to camera names and values as the images.image: For ImageType.Scene, ImageType.DepthVis, ImageType.OpticalFlowVis, image data is RGB image, numpy array with dtype=np.uint8, shape=(height, width, 3)For ImageType.DepthPerspective, ImageType.DepthPlanar, ImageType.DisparityNormalized, image data is depth image in meters, numpy array with dtype=np.float32, shape=(height, width, 1)For ImageType.Segmentation, image data is array of objects’ SegmentationIDs, numpy array with dtype=np.uint8, shape=(height, width, 1)For ImageType.SurfaceNormals, image data is array of unit vector (normalized length), numpy array with dtype=np.float32, shape=(height, width, 3)For ImageType.OpticalFlow, image data is array of displacements in x,y directions, numpy array with dtype=np.float32, shape=(height, width, 2)For ImageType.Infrared, image data is array of infrared values (object’s SegmentationID for now), numpy array with dtype=np.uint8, shape=(height, width, 1)camera parameters: width (int): image widthheight (int): image heightfov (float): camera field of view in degreescamera_position (List[float]): camera position (x,y,z) in airgen (NED frame)camera_orientation_euler_pry (List[float]): camera orientation (pitch, roll, yaw) in degreescamera_orientation_quat_wxyz (List[float]): camera orientation (w,x,y,z) in quaternion format

Mechanics

cameracoord2worldcoord

Given AirGen camera parameters, transform a camera coordinate back to AirGen world coordinates.

List[float]

A coordinate (x, y, z).

dict

Camera parameters.

np.ndarray

homo_coord_to_nonhome_coord

Turn homogeneous coordinates into non-homogeneous coordinates by factoring out the last dimension.

ivy.Array

An array of shape (..., n).

ivy.Array

An array of shape (..., n-1).

imagecoord2direction

Given camera parameters (position, pose, and FOV) and pixel coordinates, return the 3D direction of the pixel relative to the camera.

Tuple[float, float]

Coordinate of the pixel in the image in xy format.

Dict[str, Any]

Camera parameters.

Tuple[float, float, float]

Normalized unit directional vector (x, y, z).

imagecoord2orientation

Given camera parameters (position, pose, and FOV) and pixel coordinates, return the 3D direction of the pixel relative to the camera, represented in yaw and pitch (absolute degrees).

Tuple[float, float]

Coordinate of the pixel in the image in xy format.

Dict[str, Any]

Camera parameters.

Tuple[float, float, float]

Target pitch, roll, yaw in radians.

imagecoord2pose

Convert pixel coordinates to 3D coordinates.

Tuple[float, float]

Coordinate of the pixel in the image in xy format.

float

Depth of the point.

Dict[str, Any]

Camera parameters.

Tuple[float, float, float]

Target coordinate in (x, y, z) and orientation (pitch, roll, yaw) in radians.

pose2vector

Transform a target direction represented in (pitch, roll, yaw) in radians into a directional vector.

float

Pitch angle in radians.

float

Roll angle in radians.

float

Yaw angle in radians.

Tuple[float, float, float]

Unit directional vector (x, y, z).

quat_wxyz_to_xyzw

Transform a quaternion (represented by a list of floats) from wxyz format to xyzw format.

List[float]

A quaternion in wxyz format.

np.ndarray

A quaternion in xyzw format.

rotate_xy

Rotate the xy component of a 3D vector by theta (in degrees) counterclockwise in the xy plane.Assume the orientation:

^ y
|
|
|______> x

np.ndarray

A vector of shape (3,).

float

Rotation angle in degrees.

np.ndarray

Rotated vector of shape (3,).

to_eularian_angles

Transform a quaternion into Euler angles.

Quaternionr

A quaternion in wxyz format.

Tuple[float, float, float]

Euler angles (pitch, roll, yaw) in radians.

vec2eulerangles

Transform an AirGen directional vector into Euler angles.

np.ndarray

A directional vector of shape (N, 3).

np.ndarray

np.ndarray: Euler angles of shape (N, 3) where each row is (pitch, roll, yaw) in degrees.

Vision

The vision subsection of AirGen utilities provides essential functions for handling camera parameters and generating point clouds from depth images.

build_camera

Given AirGen camera parameters, this function builds the inverse extrinsic matrix and the camera intrinsic matrix.

dict

AirGen camera parameters.

Tuple[ivy.Array, ivy.Array]

Tuple of: Inverse Extrinsic Matrix, Intrinsic Matrix

build_camera_intrinsic

Constructs the camera intrinsic matrix using AirGen camera parameters.

dict

AirGen camera parameters.

ivy.Array

build_camera_inv_extrinsic

Generates the inverse extrinsic matrix for the camera based on its parameters.

dict

AirGen camera parameters.

ivy.Array

Inverse Extrinsic Matrix

camera_unproject_depth

Generates a point cloud from a depth image (Depth Perspective).

np.ndarray

Depth image of shape (H, W, 1).

ivy.Array

Inverse of camera extrinsic matrix.

ivy.Array

Inverse of camera intrinsic matrix.

np.ndarray

Point cloud of shape (N, 3).

depth2pointcloud

Generates a point cloud from a depth image.

np.ndarray

Depth image of shape (H, W, 1).

dict

Camera parameters containing field of view, height, width, and pose.

Optional[np.ndarray]

Boolean mask where 1 indicates the object of interest (H, W, 1). Defaults to None.

np.ndarray

Point cloud in AirGen world coordinates of shape (N, 3).

Trajectory

The trajectory subsection of AirGen utilities focuses on waypoint-based path planning and trajectory generation.

Waypoint

Stores a waypoint with positional and timing information.

Position in the world frame.

Euler angle of the waypoint.

Time at which the waypoint is to be reached.

calc_time

Calculates a fitting time for the trajectory between two waypoints.

Waypoint

Starting waypoint.

Waypoint

Ending waypoint.

fit_min_snap_trajectory

Starting point for trajectory generation by transforming an array of arrays into an array of waypoints.

array of arrays

Input waypoints.

bool

Indicates whether to calculate joint operations. Defaults to True.

joint

Calculates a trajectory using joint operations.

array of Waypoints

Waypoints to be connected by joint operations.

separate

Calculates a trajectory using separate operations.

array of Waypoints

Waypoints to be processed individually.

Visualize

The visualize subsection of AirGen utilities provides methods to log camera data and images from AirGen to Rerun for real-time visualization.

rr_camera_capture_logger

Given a view space name (entity_path in Rerun), returns a function that logs camera data (pose and capture in AirGen’s 3D world coordinate) to that view space.

str

Name of the view space (entity_path) in Rerun.

A callable function that logs camera data to the specified view space.

Usage:

trajectory_logger = rr_camera_capture_logger("airgen_world")
trajectory_logger(image, camera_params, camera_name)

rr_log_airgen_image

Logs images from AirGen to Rerun.

str

Path to the entity in Rerun.

ImageType

Type of image to log.

np.ndarray

Image data.

str, optional

Name of the image (default is an empty string).

If image_type is not supported.

​General

​get_public_fields

​list_to_2d_float_array

​string_to_float_array

​string_to_uint8_array

​to_dict

​to_str

​vector3d2list

​wait_key

​write_file

​Sensor

​imagetype2request

​responses2images

​Mechanics

​cameracoord2worldcoord

​homo_coord_to_nonhome_coord

​imagecoord2direction

​imagecoord2orientation

​imagecoord2pose

​pose2vector

​quat_wxyz_to_xyzw

​rotate_xy

​to_eularian_angles

​vec2eulerangles

​Vision

​build_camera

​build_camera_intrinsic

​build_camera_inv_extrinsic

​camera_unproject_depth

​depth2pointcloud

​Trajectory

​Waypoint

​calc_time

​fit_min_snap_trajectory

​joint

​separate

​Visualize

​rr_camera_capture_logger

​rr_log_airgen_image

General

get_public_fields

list_to_2d_float_array

string_to_float_array

string_to_uint8_array

to_dict

to_str

vector3d2list

wait_key

write_file

Sensor

imagetype2request

responses2images

Mechanics

cameracoord2worldcoord

homo_coord_to_nonhome_coord

imagecoord2direction

imagecoord2orientation

imagecoord2pose

pose2vector

quat_wxyz_to_xyzw

rotate_xy

to_eularian_angles

vec2eulerangles

Vision

build_camera

build_camera_intrinsic

build_camera_inv_extrinsic

camera_unproject_depth

depth2pointcloud

Trajectory

Waypoint

calc_time

fit_min_snap_trajectory

joint

separate

Visualize

rr_camera_capture_logger

rr_log_airgen_image