Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

rostopic echo /ground_plane returning [nan, nan, nan] for ground plane estimation launch file #229

Open
marcuskohtongjie95 opened this issue Jan 5, 2020 · 4 comments

Comments

@marcuskohtongjie95
Copy link

Hi all, I am running the upper_body_detector and ground_plane_estimated in simulation only (gazebo/rviz) and have installed the real_sense_gazebo_plugin (r200 camera).
I remapped the topics of ground_plane_estimation.launch to subscribe the the r200 camera as follow:

<arg name="camera_namespace" default="/r200/camera" /> <arg name="depth_image" default="/depth/image_raw" /> <arg name="camera_info_rgb" default="/color/camera_info" /> <arg name="ground_plane" default="/ground_plane" />

However when i run the ground_plane_estimated.launch and i rostopic echo /ground_plane I get the following msgs

header:
seq: 180
stamp:
secs: 4107
nsecs: 219000000
frame_id: "color"
n: [nan, nan, nan]
d: inf

header:
seq: 181
stamp:
secs: 4107
nsecs: 888000000
frame_id: "color"
n: [nan, nan, nan]
d: nan

Running the upper_body_detector image works fine, meaning that i am able to visualise the topic /upper_body_detector/image but however i am unable to get a detection (either by the bounding box in rqt_image or marker array in rviz), thus i think that the lack of detection is due to the /ground_plane topic not publishing correctly.
I initially thought the problem was because in gazebo, the depth camera is unable to capture the invisible(?) ground plane so i placed a flooring (actually a piece of wall laid flat on the floor) below the robot but that doesnt seem to solve the issue.
Can anyone help me out with this i would be very grateful thanks!
Screenshot from 2020-01-05 20-40-56

@FooKii
Copy link

FooKii commented Mar 19, 2020

hey, im also using gazebo. By running people_tracker_standalone.launch i can successully have human identified by upper-body-detection (marker seen in rviz). maybe it can help u somehow.
I do have problem when running people_tracker_robot.launch where i can have result from leg_detector but nothing from upper-body-detector.

@marcuskohtongjie95
Copy link
Author

hi thanks for the reply, i couldn't solve this issue so I changed to the darket_yolo package for human detection instead and its working great for me

@FooKii
Copy link

FooKii commented Apr 11, 2020

hi, it's me again cause i ran into the same problem.
I guess the reason is that your camera's angle or height make it impossible to see the "ground".
The solution is to switch to fixed_groundplane, but i'm guessing u then have the problem caused by "PTU" (e.x. i dont have a ptu integrated), this then involve the issue #177 which is still open.

I fixed by modifying fixed_gp.cpp (modified ugly but it works).

If you still need this, let me know, i try to make a pull request.

OR: if your robot is static or u just want it to work, a quick fix will be set /ptu_state to /joint_states in ground_plane_fixed.launch

@marcuskohtongjie95
Copy link
Author

Hi there!
My camera (RealSense r200) is placed 1 meter above the ground and sees the ground just fine. I also do not have a PTU integrated as well and my robot is moving so i don't think I can apply the quick fix that you mentioned.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants