Hello,
We’re trying to obtain 3D bounding boxes in ROS2. While everything appears correct from the viewport perspective, upon utilizing the ROS2 camera helper, the positions are aligned with the world frame and shape is inaccurate.
From Viewport:
From Rviz:
bbox ROS2 message:
header:
stamp:
sec: 0
nanosec: 83333337
frame_id: camera_tof_nfov
detections:
- header:
stamp:
sec: 0
nanosec: 0
frame_id: ‘’
results:- hypothesis:
class_id: ‘0’
score: 1.0
pose:
pose:
position:
x: 0.0
y: 0.0
z: 0.0
orientation:
x: 0.0
y: 0.0
z: 0.0
w: 1.0
covariance:- 0.0
…
bbox:
center:
position:
x: 8.0
y: 8.817369234748185e-08
z: -9.685754776000977e-07
orientation:
x: 8.876164951324077e-07
y: 2.66227889164568e-07
z: 1.4337469826566205e-07
w: 0.9999999999995604
size:
x: 1.1871184595158764
y: 1.187118459517564
z: 1.1871184595177127
id: ‘’
- 0.0
- hypothesis:
- header:
stamp:
sec: 0
nanosec: 0
frame_id: ‘’
results:- hypothesis:
class_id: ‘1’
score: 1.0
pose:
pose:
position:
x: 0.0
y: 0.0
z: 0.0
orientation:
x: 0.0
y: 0.0
z: 0.0
w: 1.0
covariance:- 0.0
…
bbox:
center:
position:
x: 8.0
y: 3.0
z: 0.0
orientation:
x: 4.440892098402098e-06
y: 4.440892098402098e-06
z: 0.9999999999778133
w: 2.220446049309721e-06
size:
x: -267315.3285292032
y: -267315.3285292032
z: -267315.3285292032
- 0.0
- hypothesis: