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

New package to advertise virtual RGB-D camera #1884

Closed
wants to merge 1 commit into from

Conversation

wkentaro
Copy link
Member

@wkentaro wkentaro commented Sep 23, 2016

Suggested in ros-perception/image_common#54.

it publishes:

  • camera_info with calibration_file.
  • image with image_file.
  • points with pcd_file.

@wkentaro wkentaro changed the title New package to advertise virtual camera New package to advertise virtual RGBD camera Sep 23, 2016
@wkentaro wkentaro force-pushed the virtual-camera-driver branch 3 times, most recently from dfeda0e to 7b51661 Compare September 23, 2016 17:17
@wkentaro wkentaro changed the title New package to advertise virtual RGBD camera New package to advertise virtual RGB-D camera Sep 23, 2016
@wkentaro wkentaro force-pushed the virtual-camera-driver branch 13 times, most recently from c4fc89b to 93303f6 Compare September 23, 2016 19:12
@k-okada
Copy link
Member

k-okada commented Sep 24, 2016

we already have package to publish virtual camera info at ->
https://github.com/jsk-ros-pkg/jsk_recognition/blob/master/jsk_perception/src/virtual_camera_mono.cpp

◉ Kei Okada

2016-09-24 2:13 GMT+09:00 Kentaro Wada [email protected]:

it publishes:

  • camera_info with calibration_file.
  • image with image_file.
  • points with pcd_file.

You can view, comment on, or merge this pull request online at:

#1884
Commit Summary

  • New package to advertise virtual camera

File Changes

Patch Links:


You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
#1884, or mute the
thread
https://github.com/notifications/unsubscribe-auth/AAeG3Nlq7QmGFchO_R8yyBN7d7psIaQLks5qtAicgaJpZM4KFMpe
.

@wkentaro
Copy link
Member Author

What do u mean?
It is not a package but a node, and it does not have feature to publish camera info.

@wkentaro wkentaro force-pushed the virtual-camera-driver branch 4 times, most recently from 3365243 to 4b4c3dc Compare September 25, 2016 09:48
@k-okada
Copy link
Member

k-okada commented Sep 26, 2016

What do u mean?

did you know this before you create new package,

It is not a package but a node, and it does not have feature to publish camera info.

This node does, and used in elevator demo @furushchev

I do not say you need to used this, and understand that your new
package seems better then this one, but adding new package increase
maintenance work and need to be very careful, at least please send
pull request to image_pipeline

$ rosnode  info virtual_camera_mono
--------------------------------------------------------------------------------
Node [/virtual_camera_mono]
Publications:
 * /virtual_camera_mono/image/compressed/parameter_descriptions
[dynamic_reconfigure/ConfigDescription]
 * /virtual_camera_mono/image/theora/parameter_descriptions
[dynamic_reconfigure/ConfigDescription]
 * /virtual_camera_mono/image/theora/parameter_updates
[dynamic_reconfigure/Config]
 * /virtual_camera_mono/image/compressed [sensor_msgs/CompressedImage]
 * /virtual_camera_mono/image [sensor_msgs/Image]
 * /rosout [rosgraph_msgs/Log]
 * /virtual_camera_mono/image/compressedDepth/parameter_descriptions
[dynamic_reconfigure/ConfigDescription]
 * /tf [tf2_msgs/TFMessage]
 * /virtual_camera_mono/camera_info [sensor_msgs/CameraInfo]
 * /virtual_camera_mono/image/compressedDepth [sensor_msgs/CompressedImage]
 * /virtual_camera_mono/image/compressedDepth/parameter_updates
[dynamic_reconfigure/Config]
 * /virtual_camera_mono/image/compressed/parameter_updates
[dynamic_reconfigure/Config]
 * /virtual_camera_mono/image/theora [theora_image_transport/Packet]

◉ Kei Okada

@wkentaro
Copy link
Member Author

wkentaro commented Sep 26, 2016

it seems that it just transports camera_info, and does not publish using calibration_file?

@wkentaro
Copy link
Member Author

Actually, this is a package which will be added to ros-drivers organization, so putting in image_pipeline seems unreasonable.

What kind of process do you need to do when a new package is added?

@k-okada
Copy link
Member

k-okada commented Sep 26, 2016

it seems that it just transports camera_info, and doea not publishes
using calibration_file?

yes, yo're correct

Actually, this is a package which will be added to ros-drivers
organization, so putting in image_pipeline seems unreasonable.

for me this seems 3d version of
https://github.com/ros-perception/image_pipeline/tree/indigo/image_publisher

◉ Kei Okada

2016-09-26 12:16 GMT+09:00 Kentaro Wada [email protected]:

Actually, this is a package which will be added to ros-drivers
organization, so putting in image_pipeline seems unreasonable.

What kind of process do you need to do when a new package is added?


You are receiving this because you commented.
Reply to this email directly, view it on GitHub
#1884 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/AAeG3GhndBgbYLi8xWu084d5RQblDz6jks5qtzkpgaJpZM4KFMpe
.

@wkentaro
Copy link
Member Author

for me this seems 3d version of
https://github.com/ros-perception/image_pipeline/tree/indigo/image_publisher

Yeah, I agree. That's why I firstly sent PR to image_common: ros-perception/image_common#54.
But what jack-oquin says seems reasonable, and virtual_camera/driver is better name than image_publisher/image_and_pcd_publisher.

@k-okada
Copy link
Member

k-okada commented Sep 28, 2016

ok, so to send this package to ros-drivers, we need to create package which
does not depends on the jsk_topic_tools, so that they can easily copy into
the ros-drivers repository.

◉ Kei Okada

2016-09-27 1:18 GMT+09:00 Kentaro Wada [email protected]:

for me this seems 3d version of
https://github.com/ros-perception/image_pipeline/
tree/indigo/image_publisher

Yeah, I agree. That's why I firstly sent PR to image_common:
ros-perception/image_common#54
ros-perception/image_common#54.
But what jack-oquin says seems reasonable, and virtual_camera/driver is
better name than image_publisher/image_and_pcd_publisher.


You are receiving this because you commented.
Reply to this email directly, view it on GitHub
#1884 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/AAeG3LKMKd8iMl2FTlCwgpf_MiXK4_Bgks5qt_BjgaJpZM4KFMpe
.

@wkentaro wkentaro force-pushed the virtual-camera-driver branch from a4e724d to 1c02534 Compare September 28, 2016 11:04
@k-okada
Copy link
Member

k-okada commented Sep 28, 2016

you'd better to create node to get accepted at ros-drives, I think.

@k-okada
Copy link
Member

k-okada commented Sep 28, 2016

I think there is a sig for ros-drivers and you'd better to start discussion there.

@wkentaro
Copy link
Member Author

I removed the dependency on jsk_topic_tools and jsk_tools.

@wkentaro
Copy link
Member Author

wkentaro commented Sep 28, 2016

you'd better to create node to get accepted at ros-drives, I think.

You mean in addition to the nodelet?

I think there is a sig for ros-drivers and you'd better to start discussion there.

Ok, I will.

@k-okada
Copy link
Member

k-okada commented Sep 28, 2016

yes, i far as i understand, you removed jsk_topic_tools and also remove
node executable, is it correct?

◉ Kei Okada

2016-09-28 20:20 GMT+09:00 Kentaro Wada [email protected]:

you'd better to create node to get accepted at ros-drives, I think.

You mean in addition to the nodelet?


You are receiving this because you commented.
Reply to this email directly, view it on GitHub
#1884 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/AAeG3BGZiXYkVj0poDscPU1vf20B_dJTks5quk2GgaJpZM4KFMpe
.

@wkentaro
Copy link
Member Author

node executable, is it correct?

Yes, you're right.

@wkentaro
Copy link
Member Author

@wkentaro wkentaro changed the title New package to advertise virtual RGB-D camera [WIP] New package to advertise virtual RGB-D camera Sep 29, 2016
@wkentaro wkentaro force-pushed the virtual-camera-driver branch 2 times, most recently from 9f49db4 to 7295457 Compare October 10, 2016 00:31
it publishes:

- camera_info with calibration_file.
- image with image_file.
- points with pcd_file.
@wkentaro wkentaro force-pushed the virtual-camera-driver branch from 7295457 to 8ddacf4 Compare October 10, 2016 00:31
@wkentaro wkentaro changed the title [WIP] New package to advertise virtual RGB-D camera New package to advertise virtual RGB-D camera Oct 10, 2016
@wkentaro
Copy link
Member Author

yes, i far as i understand, you removed jsk_topic_tools and also remove
node executable, is it correct?

Seems adding additional node executable is not recommended by OSRF maintainer ;)
ros/nodelet_core#47 (comment)

@k-okada
Copy link
Member

k-okada commented Oct 16, 2016

I think this is not a problem of "executable node", but a concern of
"maintenance cost"

◉ Kei Okada

2016-10-15 4:08 GMT+09:00 Kentaro Wada [email protected]:

yes, i far as i understand, you removed jsk_topic_tools and also remove
node executable, is it correct?

Seems adding additional node executable is not recommended by OSRF
maintainer ;)
ros/nodelet_core#47 (comment)
ros/nodelet_core#47 (comment)


You are receiving this because you commented.
Reply to this email directly, view it on GitHub
#1884 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/AAeG3DsG1HOyp9sxlMzHVrQioVOTJKUtks5qz9MUgaJpZM4KFMpe
.

@wkentaro
Copy link
Member Author

wkentaro commented Feb 5, 2017

I think this package solves our problem on CI I reported here: #1997
Because playing rosbag on Travis/Jenkins seems unstable: #1998 (comment)

What do you think? @k-okada @mmurooka @iory

@k-okada
Copy link
Member

k-okada commented Feb 5, 2017 via email

@wkentaro
Copy link
Member Author

wkentaro commented Feb 5, 2017

I see. I will go that direction.

@wkentaro
Copy link
Member Author

Maybe loading depth image to publish by OpenCV is a bit difficult, and rosbag is better for that purpose.
The problem of rosbag is its difficulty of handling --loop.
I think it can be easily resolved by rewriting header stamp with Python, and can be resolved in rosbag play with some options in the future.
So I'm closing this PR for now.

@wkentaro wkentaro closed this Mar 22, 2017
@wkentaro wkentaro deleted the virtual-camera-driver branch March 22, 2017 14:24
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants