ObjectData Struct Reference

Object data structure directly from the SDK. Represents a single object detection. More...

Attributes

int id
 Object identification number, used as a reference when tracking the object through the frames. More...
 
string uniqueObjectId
 Unique ID to help identify and track AI detections. Can be either generated externally, or using ZEDCamera.generateUniqueId() or left empty More...
 
int rawLabel
 Object label, forwarded from CustomBoxObjects when using DETECTION_MODEL.CUSTOM_BOX_OBJECTS More...
 
sl.OBJECT_CLASS label
 Object category. Identify the object type. More...
 
sl.OBJECT_SUBCLASS sublabel
 Object subclass. More...
 
sl.OBJECT_TRACKING_STATE objectTrackingState
 Defines the object tracking state. More...
 
sl.OBJECT_ACTION_STATE actionState
 Defines the object action state. More...
 
float confidence
 Defines the detection confidence value of the object. A lower confidence value means the object might not be localized perfectly or the label (OBJECT_CLASS) is uncertain. More...
 
System.IntPtr mask
 Defines for the bounding_box_2d the pixels which really belong to the object (set to 255) and those of the background (set to 0). More...
 
Vector2[] boundingBox2D
 Image data. Note that Y in these values is relative from the top of the image. If using this raw value, subtract Y from the image height to get the height relative to the bottom. More...
 
Vector3 position
 Defines the object 3D centroid. More...
 
Vector3 headPosition
 3D head centroid. More...
 
Vector3 velocity
 Defines the object 3D velocity. More...
 
Vector3[] boundingBox
 The 3D space bounding box. given as array of vertices More...
 
Vector3[] headBoundingBox
 bounds the head with eight 3D points. More...
 
Vector2[] keypoints2D
 A set of useful points representing the human body, expressed in 2D. We use a classic 18 points representation, the points semantic and order is given by BODY_PARTS. More...
 
Vector3[] keypoints
 A set of useful points representing the human body, expressed in 3D. We use a classic 18 points representation, the points semantic and order is given by BODY_PARTS. More...
 
float[] positionCovariance
 Full covariance matrix for position (3x3). Only 6 values are necessary [p0, p1, p2] [p1, p3, p4] [p2, p4, p5] More...
 
float[] keypointConfidence
 Per keypoint detection confidence, can not be lower than the ObjectDetectionRuntimeParameters.detectionConfidenceThreshold. Not available with DETECTION_MODEL.MULTI_CLASS_BOX. in some cases, eg. body partially out of the image or missing depth data, some keypoint can not be detected, they will have non finite values. More...
 
Vector3[] localPositionPerJoint
 Global position per joint in the coordinate frame of the requested skeleton format. More...
 
Quaternion[] localOrientationPerJoint
 Local orientation per joint in the coordinate frame of the requested skeleton format. The orientation is represented by a quaternion. More...
 
Quaternion globalRootOrientation
 Global root position. More...
 

Detailed Description

Object data structure directly from the SDK. Represents a single object detection.

Variables

◆ id

int id

Object identification number, used as a reference when tracking the object through the frames.

◆ uniqueObjectId

string uniqueObjectId

Unique ID to help identify and track AI detections. Can be either generated externally, or using ZEDCamera.generateUniqueId() or left empty

◆ rawLabel

int rawLabel

Object label, forwarded from CustomBoxObjects when using DETECTION_MODEL.CUSTOM_BOX_OBJECTS

◆ label

Object category. Identify the object type.

◆ sublabel

Object subclass.

◆ objectTrackingState

sl.OBJECT_TRACKING_STATE objectTrackingState

Defines the object tracking state.

◆ actionState

Defines the object action state.

◆ confidence

float confidence

Defines the detection confidence value of the object. A lower confidence value means the object might not be localized perfectly or the label (OBJECT_CLASS) is uncertain.

◆ mask

System.IntPtr mask

Defines for the bounding_box_2d the pixels which really belong to the object (set to 255) and those of the background (set to 0).

◆ boundingBox2D

Vector2 [] boundingBox2D

Image data. Note that Y in these values is relative from the top of the image. If using this raw value, subtract Y from the image height to get the height relative to the bottom.

0 ----— 1 | obj | 3-----— 2

◆ position

Vector3 position

Defines the object 3D centroid.

◆ headPosition

Vector3 headPosition

3D head centroid.

◆ velocity

Vector3 velocity

Defines the object 3D velocity.

◆ boundingBox

Vector3 [] boundingBox

The 3D space bounding box. given as array of vertices

1 ------—2 /| /| 0 |-----—3 | | | | | | 5-----—|-6 |/ |/ 4 ------—7

◆ headBoundingBox

Vector3 [] headBoundingBox

bounds the head with eight 3D points.

◆ keypoints2D

Vector2 [] keypoints2D

A set of useful points representing the human body, expressed in 2D. We use a classic 18 points representation, the points semantic and order is given by BODY_PARTS.

◆ keypoints

Vector3 [] keypoints

A set of useful points representing the human body, expressed in 3D. We use a classic 18 points representation, the points semantic and order is given by BODY_PARTS.

◆ positionCovariance

float [] positionCovariance

Full covariance matrix for position (3x3). Only 6 values are necessary [p0, p1, p2] [p1, p3, p4] [p2, p4, p5]

◆ keypointConfidence

float [] keypointConfidence

Per keypoint detection confidence, can not be lower than the ObjectDetectionRuntimeParameters.detectionConfidenceThreshold. Not available with DETECTION_MODEL.MULTI_CLASS_BOX. in some cases, eg. body partially out of the image or missing depth data, some keypoint can not be detected, they will have non finite values.

◆ localPositionPerJoint

Vector3 [] localPositionPerJoint

Global position per joint in the coordinate frame of the requested skeleton format.

◆ localOrientationPerJoint

Quaternion [] localOrientationPerJoint

Local orientation per joint in the coordinate frame of the requested skeleton format. The orientation is represented by a quaternion.

◆ globalRootOrientation

Quaternion globalRootOrientation

Global root position.