Back to Jetson Inference

Jetson Inference: poseNet::ObjectPose Struct Reference

docs/html/structposeNet_1_1ObjectPose.html

latest5.1 KB
Original Source

| | Jetson Inference

DNN Vision Library |

Classes | Public Member Functions | Public Attributes | List of all members

poseNet::ObjectPose Struct Reference

The pose of an object, composed of links between keypoints. More...

#include <poseNet.h>

|

Classes

| | struct | Keypoint | | | A keypoint or joint in the topology. More...
| | |

|

Public Member Functions

| | int | FindKeypoint (uint32_t id) const | | | Find a link index by two keypoint ID's, or return -1 if not found. More...
| | | | int | FindLink (uint32_t a, uint32_t b) const | | |

|

Public Attributes

| | uint32_t | ID | | | Object ID in the image frame, starting with 0. More...
| | | | float | Left | | | Bounding box left, as determined by the left-most keypoint in the pose. More...
| | | | float | Right | | | Bounding box right, as determined by the right-most keypoint in the pose. More...
| | | | float | Top | | | Bounding box top, as determined by the top-most keypoint in the pose. More...
| | | | float | Bottom | | | Bounding box bottom, as determined by the bottom-most keypoint in the pose. More...
| | | | std::vector< Keypoint > | Keypoints | | | List of keypoints in the object, which contain the keypoint ID and x/y coordinates. More...
| | | | std::vector< std::array< uint32_t, 2 > > | Links | | | List of links in the object. More...
| | |

Detailed Description

The pose of an object, composed of links between keypoints.

Each image can have multiple objects detected per frame.

Member Function Documentation

FindKeypoint()

|

| int poseNet::ObjectPose::FindKeypoint | ( | uint32_t | id | ) | const |

| inline |

Find a link index by two keypoint ID's, or return -1 if not found.

This returns an index into the Links list

|

| int poseNet::ObjectPose::FindLink | ( | uint32_t | a, | | | | uint32_t | b | | | ) | | const |

| inline |

Member Data Documentation

Bottom

| float poseNet::ObjectPose::Bottom |

Bounding box bottom, as determined by the bottom-most keypoint in the pose.

ID

| uint32_t poseNet::ObjectPose::ID |

Object ID in the image frame, starting with 0.

Keypoints

| std::vector<Keypoint> poseNet::ObjectPose::Keypoints |

List of keypoints in the object, which contain the keypoint ID and x/y coordinates.

Left

| float poseNet::ObjectPose::Left |

Bounding box left, as determined by the left-most keypoint in the pose.

| std::vector<std::array<uint32_t, 2> > poseNet::ObjectPose::Links |

List of links in the object.

Each link has two keypoint indexes into the Keypoints list Find a keypoint index by it's ID, or return -1 if not found. This returns an index into the Keypoints list

Right

| float poseNet::ObjectPose::Right |

Bounding box right, as determined by the right-most keypoint in the pose.

Top

| float poseNet::ObjectPose::Top |

Bounding box top, as determined by the top-most keypoint in the pose.


The documentation for this struct was generated from the following file: