Frame_id + Publishing depth information about ros_openpose HOT 6 CLOSED

@aabsz

Thank you very much for your warm message. I appreciate it. I think the credit goes to our wonderful contributors.

... getting an almost fine skeleton in Rviz ... the output of Rviz is not perfect,...

I would suggest sharing a screenshot of your RViz, please.

I think it might be due to not using a correct "frame_id".

We have covered this topic in the FAQ. May I request to look at the same, please? Furthermore, check out the existing configs for various cameras inside the launch subdirectory. Nevertheless, please feel free to write back if it does not answer your question. I would love to improve the documentation based on your feedback.

How can I get the depth data of each joint along with its 2D position too?

I am sorry, do you mean each joint's 3D coordinate (x, y, and z values)?

On a side note, once you finish this task, may I ask you to share your config_xtion.launch? This valuable information will be made available to everyone.

from ros_openpose.

Dear Rajivo,

Thank you for the prompt reply.
I'll share the RVIZ screen ASAP.

I had already checked the FAQ and all config.launch files of other sensors, but I still don't know what it is exactly for Xtion. I think I should search more about it.

Regarding the 3D positions, yes, exactly. I need access to each joint's 3D position data and feed these data as input to a DL model. Currently, the 2D positions are published, but I also need access to the depth information. Has this already been included in the code, or do I have to do it myself? Do you have any suggestions on how to do it?

I just made some minor changes to the config.launch code to make it compatible with Xtion. However, I would be happy to share the final code with you so that everyone can use it :-)

from ros_openpose.

Thanks @aabsz for the update

Regarding the 3D positions, yes, exactly.

This is available already. Right now, in the terminal, you can see the pixels only just to keep the terminal cleaner. Please see the current implementation below:

I need access to each joint's 3D position data and feed these data as input to a DL

You should subscribe to the "/frame" topic to receive the 3D position data.

I just made some minor changes to the config.launch code to make it compatible with Xtion. However, I would be happy to share the final code with you so that everyone can use it :-)

Please share necessary files and documentation. If you are not sure about GitHub, share via email, please. My email is mentioned here.

from ros_openpose.

Dear ravijo,

Sorry for the long delay. I figured out everything regarding the frame_id for the Xtion RGB-D camera, which is now solved. I will share the corresponding Config file for this camera.
Regarding the depth information of each joint, I can not get the depth information from the depth sensor yet. Is it possible to guide me more on how to get this data?
I'm getting each joint's pixel information on the terminal. I want to get the depth information too.

Many thanks

from ros_openpose.

@aabsz

I can not get the depth information from the depth sensor yet

I think you mean depth sensor ros_openpose.

I want to get the depth information too.

You should subscribe to the "/frame" topic to receive the 3D position data.

I suggested you to get the 3D position data from "/frame" topic. Just execute the following in terminal:

rostopic echo /frame

Or if you want to do it via code, change this line as following

text = [bodyPart.point for person in msg.persons for bodyPart in person.bodyParts]

from ros_openpose.

Closing this issue because of no further comments. However, feel free to comment if the problem still exists.

from ros_openpose.