You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
hello, I have a question about the way to get the alpha using the heading
in your examples, the groundtruth_extraction.py file, the code is
126 pos = (label.box.center_x,label.box.center_y,label.box.center_z,1)
127 pos = np.matmul(camera_extrinsic_inv, pos)
130 alpha = label.box.heading - math.atan2(pos[2],pos[0])
the pos[2] and pos[0] is center_z and center_x in the camera frame
however in the camera frame of Waymo Dataset, z points up,
so I think maybe the code change to alpha = label.box.heading - math.atan2(pos[1],pos[0])
Is it right or wrong?
and another question: label.box.heading is in the vehicle frame , math.atan2(pos[2],pos[0]) is in the camera frame.
so the alpha is camera frame or not?
The text was updated successfully, but these errors were encountered:
Thanks for reporting, yes, the code sounds wrong. I used the heading but didn’t rely on the alpha so I might not have noticed the wrong value. I haven’t used the dataset in a while so if you have tested the fix, feel free to do a pull request. Otherwise I can have a look when I have got some spare time.
hello, I have a question about the way to get the alpha using the heading
in your examples, the groundtruth_extraction.py file, the code is
126
pos = (label.box.center_x,label.box.center_y,label.box.center_z,1)
127
pos = np.matmul(camera_extrinsic_inv, pos)
130
alpha = label.box.heading - math.atan2(pos[2],pos[0])
the pos[2] and pos[0] is center_z and center_x in the camera frame
however in the camera frame of Waymo Dataset, z points up,
so I think maybe the code change to
alpha = label.box.heading - math.atan2(pos[1],pos[0])
Is it right or wrong?
and another question:
label.box.heading
is in the vehicle frame ,math.atan2(pos[2],pos[0])
is in the camera frame.so the alpha is camera frame or not?
The text was updated successfully, but these errors were encountered: