Image not viewing on rqt

I’m trying to use the turbtlebot3 tutorial for lane detection.

I am trying to view an image there but for
some reason it is not showing.

I am at this step:
step 8.2.3
and whenever I try to see what is the image being published to topic: /camera/image_projected_compensated
I see nothing…
Has anyone tried to use this on the magni?

What I don’t know what’s the issue with running it…
Any help would be appreciated

EDIT: This was fixed, I was initially running the calibration files on the workstation. However, when I ran all the commands on the magni, except for the rqt (because that needs a screen interface), ran fine.

NOTE: Keep in mind when using other open source code projects to see how the nature of the robot setup is. In the case of this tutorial, the turtlebot here is setup as the slave and the Workstation/remote PC is the Master. This isn’t the case with the magni. This could help a lot in different things

If you have not modified launch files and so on and are running raspicam node then the default image topic is: /raspicam_node/image/compressed

In order to even see this you must start the raspicam node because it is not just on automatically like the motor node is always on. See this page if you have not already:

So are you saying your nodes subscribe to raspicam image topic from above ok but cannot see them re-published to /camera/image_projected_compensated that is part of your turtlebot3 packages you run?

Sometimes: rosnode info my_node is useful to show input and output topics (subscribed and published topics).

So, I have looked at the launch files and there seems to be nothing in terms of topics that needed to be adjusted.
The raspicam node is launched by using the camera launch file, which specifies the camera, which is used by the link you have me above.
And yeah, for some reason, the image are not being published or rather not showing up to other topics…
one thing i noticed whenever I choose those topics on rqt image viewer is this:

[ERROR] [1618487421.349872734]: Tried to advertise a service that is already advertised in this node [/rqt_gui_cpp_node_4143/compressed/set_parameters]

and another thing I found was when running rqt_graph, this showed up:

does the image compensation projection here mean it is not communicating with anything else?

You do have rates going into the two nodes on lower right.
I don’t see if those functions have any output or not.
Are those two on the lower right properly doing callbacks off of their inbound image topic subscriptions? maybe they are not processing the inbound subscriptions for some reason.
Nothing shows to the right so I cannot see if they have output or not of any rate.

Okay, so I did something different. I’m not sure why it worked, but basically, what I did was instead of running the calibration launch files, and the lane detection launch file on the remote PC, I ran them on the magni itself.

What I think happened is that for some reason, because the way the turtlebot is set up, it caused that connection error? I don’t know what it is exactly.

The only explanation I could come up with is, the turtlebot system is run where the workstation the master of the robot, and the robot is the slave/host. However, in the case of the magni, it’s the magni that is the master and the workstation is the slave.

The weird thing is, there isn’t an indication as to why that happened? nothing specifically I could find in the files that would state that? and the other thing is… the other camera calibration related files ran just fine…

I am also thinking of doing a tutorial using the turtlebot tutorial here on the forum for whoever wants to use it? is there anything restrictions with regards to that?

Start a new thread with a title like ‘Autonomous driving with a Magni’ or something. If you have experimented with software that is interesting like this and got it to run on a Magni I suspect other folks would be curious to hear of your findings. This forum is to assist with questions but also to share unique experiences using the Magni platform so that others can try these things as well. Keep in mind you will get persons of assorted capabilities so be ready to explain things to persons who may not be at your level of ROS and code experience. I’ll read it for sure, it is an interesting area for robotics.

1 Like

I will do that actually, I believe it would help a lot for people who want to do their projects