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
However, if I run the robot in simulation inside Gazebo and I try to obtain the intrinsic parameters of the depth part, e.g. using rpc directly, I get the following answer:
with $w = 640$ the width and $\mathrm{fov}_h$ the horizontal field of view from the urdf.
Hence, it seems that the parameters from the configuration file are ignored - which I actually also like as otherwise they might differ from the field of view inside the urdf.
Is anybody aware if the parameters inside the [CAMERA_PARAM] are actually used or not? Should we remove them?
While looking at the configuration files used for cameras in simulation, I noticed that the one used for the depth part, i.e.,
ergocub-software/urdf/ergoCub/robots/ergoCubGazeboV1_1/model.urdf
Lines 3479 to 3481 in 001eef3
contains a section
ergocub-software/urdf/ergoCub/conf/sensors/gazebo_ergocub_rgbd_camera.ini
Lines 4 to 8 in 001eef3
with focal parameters.
However, if I run the robot in simulation inside Gazebo and I try to obtain the intrinsic parameters of the depth part, e.g. using rpc directly, I get the following answer:
which tells that the focal lengths are$f_x = f_y = 457.02216941509868775$ that are different from those from the configuration file.
Indeed, the field of view used for the depth camera inside the urdf is:
ergocub-software/urdf/ergoCub/robots/ergoCubGazeboV1_1/model.urdf
Line 3459 in 001eef3
whose corresponding focal length is:
with$w = 640$ the width and $\mathrm{fov}_h$ the horizontal field of view from the
urdf
.Hence, it seems that the parameters from the configuration file are ignored - which I actually also like as otherwise they might differ from the field of view inside the
urdf
.Is anybody aware if the parameters inside the
[CAMERA_PARAM]
are actually used or not? Should we remove them?Thank you
cc @traversaro @Nicogene
The text was updated successfully, but these errors were encountered: