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.

Sign up for GitHub

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

Jump to bottom

Ouster lidar os1 64 #4

Closed
islamtalha01 opened this issue Sep 1, 2021 · 43 comments
Closed

Ouster lidar os1 64 #4

islamtalha01 opened this issue Sep 1, 2021 · 43 comments

Comments

@islamtalha01
Copy link

Can this package be used for calibration of ouster os2 64 and Ip camera with this package? if it is possible then kindly guide me to complete the procedure I am doing as the research intern at NUST Pakistan. Kindly help me to achieve this task.
@chinitaberrio
Copy link

Just checkin the ROS driver for the Ouster lidar, I think it should work as it has the ring field in the point time, which is a requirement for the usage of the package.

@islamtalha01
Copy link
Author

islamtalha01 commented Sep 3, 2021 via email

@islamtalha01
Copy link
Author

islamtalha01 commented Sep 3, 2021 via email

@chinitaberrio
Copy link

I'm unable to see the attached files

@islamtalha01
Copy link
Author

islamtalha01 commented Sep 3, 2021 via email

@chinitaberrio
Copy link

You need to change the global frame in rviz to your lidar frame

@islamtalha01
Copy link
Author

islamtalha01 commented Sep 3, 2021 via email

@chinitaberrio
Copy link

You should be able to change the global frame in rviz, refer to this document: http://wiki.ros.org/rviz/UserGuide
Once you have launched all your drivers, open a new terminal and type rostopic list, you should be able to see all the topics published by the drivers, you can also take a look at these links: http://wiki.ros.org/rostopic http://wiki.ros.org/ROS/Tutorials/UnderstandingTopics

@islamtalha01
Copy link
Author

islamtalha01 commented Sep 3, 2021 via email

@islamtalha01
Copy link
Author

islamtalha01 commented Sep 5, 2021 via email

@chinitaberrio
Copy link

If you have any issues with your ouster driver please address them at: https://github.com/ouster-lidar/ouster_example, we can only support problems related to the camera-lidar calibration package.

@islamtalha01
Copy link
Author

islamtalha01 commented Sep 5, 2021 via email

@chinitaberrio
Copy link

Prepare a rectangular chessboard printout, preferably with odd-numbered grid dimensions (e.g. 9x7, 11x13)
Firmly attach the chessboard on a rigid, opaque, and rectangular board such that both their centres align and their edges remain parallel to one another.
Choose a suitable stand that can mount the target with little to no protruding elements from the board's edges.
Mount the target on the stand such that it is tilted at an angle of 45-60 degrees with respect to the ground plane.

You also need to adjust the experimental region for your case. The in rqt_reconfigure, feature_extraction variables correspond to the bounds of the 3D experimental region along the LiDAR's x,y,z axis:
x_min [rear], x_max [front] with range of [-6.0, 6.0]
y_min [right], y_max [left] with range of [-5.0, 5.0]
z_min [bottom], z_max [top] with range of [-2.0, 2.0]

@islamtalha01
Copy link
Author

islamtalha01 commented Sep 6, 2021 via email

@darrenjkt
Copy link
Contributor

When you run roslaunch cam_lidar_calibration run_optimiser.launch, two windows will pop up. One is Rviz, the other is the rqt_reconfigure window. The rqt_reconfigure window is where you can adjust the experimental region. That window basically sets the maximum and minimum x,y and z value such that you only allow certain points through within those bounds and filter the rest out.

@islamtalha01
Copy link
Author

islamtalha01 commented Sep 6, 2021 via email

@chinitaberrio
Copy link

Make sure the topic name is correct. Also you can refer to this readme file: https://github.com/chinitaberrio/cam_lidar_calibration

@darrenjkt
Copy link
Contributor

Before adjusting anything in the rqt_reconfigure, you should be able to see the entire pointcloud if you've configured RViz correctly

@islamtalha01
Copy link
Author

islamtalha01 commented Sep 6, 2021 via email

@islamtalha01
Copy link
Author

islamtalha01 commented Sep 6, 2021 via email

@chinitaberrio
Copy link

The first step is to have the proper setup:

Prepare a rectangular chessboard printout, preferably with odd-numbered grid dimensions (e.g. 9x7, 11x13)
Firmly attach the chessboard on a rigid, opaque, and rectangular board such that both their centres align and their edges remain parallel to one another.
Choose a suitable stand that can mount the target with little to no protruding elements from the board's edges.
Mount the target on the stand such that it is tilted at an angle of 45-60 degrees with respect to the ground plane.

The tool provided won't work if you are holding the checkerboard as it needs to be isolated from the background by adjusting the experimental region like this:

image

@darrenjkt
Copy link
Contributor

You can send me an email at jtsa5115@uni.sydney.edu.au and we can chat further.

@ygsq-123
Copy link

ygsq-123 commented Sep 8, 2021

Is the problem of adjusting the rqt_reconfigure window but the point cloud range unchanged? I also encountered the same problem. Can you tell me how to solve it, thank you.

@islamtalha01
Copy link
Author

islamtalha01 commented Sep 8, 2021 via email

@chinitaberrio
Copy link

chinitaberrio commented Sep 8, 2021 •
edited
Loading

Can you show me the point cloud topics that appear on rviz? Eg.
image

Which rostopic are you displaying on rviz?

@darrenjkt
Copy link
Contributor

darrenjkt commented Sep 8, 2021 •
edited
Loading

I've made some annotations for the rviz setup process. Let me know if this helps.
RVIZ instructions

@islamtalha01
Copy link
Author

islamtalha01 commented Sep 8, 2021 via email

@islamtalha01
Copy link
Author

islamtalha01 commented Sep 29, 2021 via email

@islamtalha01
Copy link
Author

islamtalha01 commented Sep 29, 2021 via email

@jvnsg
Copy link

jvnsg commented Jun 16, 2022

Hey @islamtalha01,
I have some similar issues like you had configuring the calibration tool for the Ouster LIDAR. Is there a way I can contact you directly to help me out?

@jogyul00
Copy link

I'm using the os1 64ch lidar, too. but I'm not able to extract the chekerboard. How did you do?? Help me, please. Just like your situation before, I adjust the rqt_reconfigure window but the point cloud range unchanged.

@TunnaBaitha
Copy link

Hey @islamtalha01 ,
I am facing the problem in Ouster LIDAR and Camera calibration. can you help me solve the problem.....

@TunnaBaitha
Copy link

did anybody get the solution of this issue? if so please help me to calibrate OS2 64 channel
Lidar and camera .....

@mumu-o
Copy link

mumu-o commented Oct 8, 2023

did anybody get the solution of this issue? if so please help me to calibrate OS2 64 channel Lidar and camera .....
Hi @TunnaBaitha
I got the same issue when using OS2 64 channel LiDAR, but I still didn't find a solution yet. Have you solved it?

@chinitaberrio
Copy link

@mumu-o, What specific problem are you experiencing?

@mumu-o
Copy link

mumu-o commented Oct 8, 2023

@mumu-o, What specific problem are you experiencing?

Hi @chinitaberrio,
Thank you for your reply.
I currently use the Ouster 64 channel LiDAR with a simple web cam for the calibration. I have modified the topics as the devices I used. However, It seems like I cannot isolate the point cloud data on the chessboard using rqt_reconfigure_para, even though I changed the config setting from rqt panel, the visualization of point cloud data didn't change. Please see the figure below.
Screenshot from 2023-10-07 22-18-23
From the rqt window, the topic that related to point cloud listed on the left is the "feature_extraction". However, from my visualization window, when I change the "Cheesboard" topic to "feature_extraction/cheesboard", there is no point cloud data shown on the right. Only when I select "ouster/point", it works. My understanding is the "feature_extraction" topic tried to convey point cloud data from the LiDAR topic ("ouster/point"), however, it didn't work from my site. So when I use rqt to change the region, the points cannot be isolated. I am not sure if my understanding is correct.
Screenshot from 2023-10-07 22-16-54
What should I do to adjust the ROI of the point cloud data?
Thank you

@chinitaberrio
Copy link

chinitaberrio commented Oct 8, 2023 •
edited
Loading

@mumu-o
Here are a few things to check first:

  1. Read the readme carefully, showing what needs to be adjusted step by step. Could you ensure you have changed all the topics and extra parameters in the config file?
    2)The only topic you must change in rviz is the camera_image and the reference frame.
    3)Are you publishing camera_image, camera_info, and point cloud in the same reference time? This package works if the topics are in sync.
  2. Have you calibrated your camera? Are the parameters correct in the camera_info topic? (without the correct intrinsic parameters, the package won't work.)

@mumu-o
Copy link

mumu-o commented Oct 9, 2023

Hi @chinitaberrio,
Thank you for the reply.
1.1), yes, I have change the topics and modified the camera information based on the result of camera calibration.
Screenshot from 2023-10-08 21-13-41
2),yes, I have changed the frame to the ouster_lidar frame, and the camera_image topic to /usb_cam/image_raw
3), I think this is the reason. I recorded a bag file using both lidar and camera topics and the figure below shows the timestamp.
Screenshot from 2023-10-08 21-11-24
Screenshot from 2023-10-08 21-21-04
2. Yes, I have finished calibrated the camera and changed the parameters based on the camera calibration result. I think the intrinsic parameters should be correct.
I think the timestamp synchronization should be the reason, but I am so confused about how to synchronize the timestamp of two devices. Could you please give me some help?
Thank you

@chinitaberrio
Copy link

chinitaberrio commented Oct 9, 2023 •
edited
Loading

@mumu-o

  1. I can't see the camera info topic in your rosbags. Intrinsic parameters are required for the Extrinsic calibration.
  2. you can check if the topics are synchronized by looking at the header timestamp, not the rosbag time.

@mumu-o
Copy link

mumu-o commented Oct 9, 2023

@chinitaberrio

  1. the intrinsic parameters are as below.
    Screenshot from 2023-10-08 21-30-20 I have modified the default value with these intrinsic values.
  2. I am confused about how to check the header timestamp. If it works if the fps of two devices don't match?
    Thank you

@chinitaberrio
Copy link

chinitaberrio commented Oct 9, 2023 •
edited
Loading

  1. In rqt_bag, select the topic with a right-click and go to view -> raw. Use the arrows in the subsequent panel to navigate. Access the 'header' dropdown, followed by the 'stamp' dropdown to view the timestamp, presented in both seconds and nanoseconds. Perform this for every topic and compare the seconds across all the ROS topics. If the difference is in the magnitude of seconds you need to synchronise your hardware, e.g. for ouster, you can use PTP.

@mumu-o
Copy link

mumu-o commented Oct 9, 2023

@chinitaberrio Thank you so much, I think this is the reason. The timestamps of my two devices are as follow.
Screenshot from 2023-10-08 21-50-00 I will work on synchronizing the timestamp.

@mumu-o
Copy link

mumu-o commented Oct 12, 2023

Hi there,
I'd like to share my solution to a potential issue that Ouster LiDAR users might encounter, as I faced a similar problem. In my setup, I used a simple webcam and the Ouster LiDAR O2.

  1. First, ensure you've launched the LiDAR, camera, and calibration topics.
  2. Change the frame (in my case, it was "os_sensor"), and the image topic (usb_cam/image_raw in my case). Note that the point cloud should be shown at this time. However, I cannot see the point cloud data.
  3. If you're still encountering problems, it may be caused by non-synchronized timestamps from different devices. In my setup, I observed timestamps as shown below:
    Timestamps
  4. To address timestamp synchronization issues, follow the guide PTP. When configuring the PTP grand master clock, ensure your Ethernet interface supports hardware or software only. If it's software-only, don't forget to change the "time_stamp" from "hardware" to "software". After completing the PTP setup, your timestamps should be synchronized, as shown below:
    Synchronized Timestamps
  5. Restart the LiDAR, camera, and calibration topics, and the package should work correctly:
    System Working
    Hope this should resolve your issue. Correct me if I am wrong.

@chinitaberrio chinitaberrio mentioned this issue Nov 8, 2023
Closed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
8 participants
@chinitaberrio @darrenjkt @jvnsg @mumu-o @ygsq-123 @islamtalha01 @jogyul00 @TunnaBaitha