cameraIMUCalibrationOptions

Options for calibrating camera and IMU

Since R2024a

Description

The cameraIMUCalibrationOptions object stores options for calibrating a transformation between a camera reference frame to an IMU reference frame for use in the estimateCameraIMUTransform function.

Creation

Syntax

options = cameraIMUCalibrationOptions

options = cameraIMUCalibrationOptions(Name=Value)

Description

options = cameraIMUCalibrationOptions creates options for calibrating a transformation between a camera to an IMU with default properties for the estimateCameraIMUTransform function.

options = cameraIMUCalibrationOptions(Name=Value) sets properties using one or more name-value arguments.

Properties

UndistortPoints — Undistort image points before calibration

1 or true (default) | 0 or false

Indication of whether the estimateCameraIMUTransform function undistorts the specified image points before calibration, specified as a logical 1 (true) or 0 (false).

InitialTransform — Initial transformation from camera frame to the IMU frame

se3() | se3 object | rigidtform3d object

Initial transformation from the camera frame to the IMU frame, specified as either a se3 object or a rigidtform3d (Image Processing Toolbox) object.

If you have a mechanical diagram for the system, you can compute this transformation. If you must recalibrate for regular maintenance, use a previously computed estimate.

SolverOptions — Solver options for calibration optimization

factorGraphSolverOptions() (default) | factorGraphSolverOptions object

Solver options for calibration optimization, specified as a factorGraphSolverOptions object.

CameraPoses — Camera poses

[] (default) | N-by-7 matrix | N-element array of se3 objects or rigidtform3d objects

Camera poses, specified as an N-by-7 matrix, as a two-column camera pose table, or as an N-element array of se3 objects or rigidtform3d (Image Processing Toolbox) objects. N is the total number of camera poses and must be equal to the length of imagePoints.

If you specify an N-by-7 matrix, each row is a 3-D pose in the form [x y z qw qx qy qz]. x, y, and z comprise the xyz position. qw, qx, qy, and qz comprise the orientation as a quaternion.

If you specify an N-element array of se3 objects or rigidtform3d objects, then all elements in the array must be of the same type. For example, if the first element contains an se3 object, all other elements must be se3 objects.

You can use the poses (Computer Vision Toolbox) object function of the imageviewset (Computer Vision Toolbox) object to get a two-column camera pose table.

If CameraPoses is empty, then the estimateCameraIMUTransform function computes the camera poses using the imagePoints argument.

CameraInformation — Camera information

eye(2) (default) | 2-by-2 matrix

Camera information, specified as a 2-by-2 matrix. The information value shows confidence in pattern point detections. Use a larger information value to specify larger confidence during calibration optimization.

ImageTime — Image time stamps

datetime.empty() | datetime object

Image time stamps, specified as a datetime object.

ImageSampleRate — Image sample rate of camera

[] | positive integer

Image sample rate of the camera, specified as a positive integer in hertz. Use this parameter when you do not have accurate image time stamps. For better estimation results, specify accurate image time stamps if available.

ShowProgress — Visualize progress in figure

"all" (default) | "undistortion" | "camera-poses" | "none"

Visualize progress in figure, specified as one of these options:

• "all" — Show the progress of undistortion and camera pose estimation.

• "undistortion — Show the progress of the undistortion.

• "camera-poses" — Show the progress of the camera pose estimation.

• "none" — Disable progress visualization.

Version History

Introduced in R2024a