SDK Calibration Data

struct ExtrinsicsReport

Structure to store the extrinsics calibration report.

Public Members

std::vector<cv::Mat> per_view_rvecs

Rotation vector from pattern to first camera for each view.

std::vector<cv::Mat> per_view_tvecs

Translation vector from pattern to first camera for each view.

std::vector<cv::Mat> per_view_errors

Re-projection error for each view of each camera.

double rmse

Root mean square error of all re-projections.

struct IntrinsicsReport

Structure to store the intrinsic calibration report.

Public Members

std::uint32_t sensor_width

Width of the sensor.

std::uint32_t sensor_height

Height of the sensor.

std::vector<bool> selected_views

Views retained for calibration.

std::vector<cv::Point3f> obj_pts

Pattern points in the pattern coordinate system.

std::vector<std::vector<cv::Point2f>> img_pts_per_view

Detections of the pattern corners in each view.

double rmse

Root mean square error of all re-projections.

std::vector<double> per_view_rms_reprojection_errors

Root mean square error of re-projections for each view.

std::vector<cv::Vec3d> tvecs

Rotation and translation from pattern to camera for each view.

struct Pattern

Structure defining the calibration pattern.

A pattern is defined by a list of 3d corners ordered by their unique ID. Besides, different patterns can have different properties that are stored as metadata.

Note

For more information about the available properties available for each pattern type, please refer to the Metavision’s documentation

Public Types

enum class Type

Enumeration of the different pattern types.

Values:

enumerator Chessboard

enumerator CirclesGrid

enumerator ActiveMarker

using Parameter = std::variant<std::uint32_t, float>

Different types of parameters that can be stored in the metadata.

Public Members

Type type

Type of the pattern.

std::map<std::uint32_t, cv::Point3f> corners_by_id

3D corners of the pattern stored by id

std::unordered_map<std::string, Parameter> metadata

Metadata of the pattern.

struct PatternDetection

Represents the detection of a pattern by a camera.

Public Members

std::vector<PointDetection> points

List of detected pattern’s corners.

cv::Mat detection_image

Image of the detection (depending on the pattern detector, can be the image used to detect the corners or an image generated for visualization purposes)

struct PointDetection

2D detection of a pattern’s 3D corner

Public Members

size_t id

Unique ID of the corner.

cv::Point2f position

Re-projection of the corner in the sensor’s plane.