-
Notifications
You must be signed in to change notification settings - Fork 47
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
'object_recognition_core detect' throws error #30
Comments
Hi, Since recent updates of ROS, ORK_core happens to run into this error. It A workaround that I use when running into this is to make sure that all the HTH. On Wed, Nov 5, 2014 at 7:52 AM, YuOhara [email protected] wrote:
|
I think all the topics that ork_detection subscribes( * /camera/depth/camera_info[sensor_msgs/CameraInfo] * /camera/rgb/image_raw [sensor_msgs/Image]* /camera/depth/image_raw [sensor_msgs/Image] * /camera/rgb/camera_info [sensor_msgs/CameraInfo] I played the program with gdb. I put the result here.
|
hi, that got fixed with the latest ecto (0.6.7) and ecto_ros (0.4.5). Sorry for the inconvenience. |
Hi, I'm also trying to follow the quick start guide to use ORK with ROS and I'm running into the same error. Running openni_launch seems to be working and the camera is publishing to the appropriate topics. rviz can access and use them just fine. I've verified that my ros-indigo-ecto package is at version 0.6.7 and ros-indigo-ecto-ros is at 0.4.5. I've tried installing and re-installing from compiled ros packages as well as compiling the ROS packages from source as detailed on the install page, with the same results. I tried again today with the new version 0.4.6 of ros-indigo-ecto-ros and I get the same result. Is there anything else which could be causing this problem? |
can you please try with the latest from ecto (the problem was on the ecto side, not ecto_ros side). It's version 0.6.8. |
Hi, For my previous tests I was following the quickstart guide by running: The second command gave the assertion failed error. I tried those scripts you mention while ecto 0.6.7 was still installed. I was unable to get consistent results - sometimes some of them would run, sometimes they would give the same error as before with the failed assert. I uninstalled 0.6.7 and attempted to build 0.6.8 from source (because it is not yet in the ROS repositories for Ubuntu) and again tried to run the scripts. One of them worked once, but all the other times it failed on a seg-fault immediately after printing the initial "Registration? 1 Sync? 0". It is possible something got messed up when I was installing from source though or from some interaction with the old version from the repositories. |
For my tests I was following the quickstart guide by running: Then I go through other tests, "rosrun object_recognition_capture orb_template -o my_textured_plane" crashs too if I launch any of those scripts in: https://github.com/plasmodic/ecto_opencv/tree/master/samples/rgbd I am also sure the following topics ( * /camera/depth/camera_info[sensor_msgs/CameraInfo] * /camera/rgb/image_raw [sensor_msgs/Image]* /camera/depth/image_raw [sensor_msgs/Image] * /camera/rgb/camera_info [sensor_msgs/CameraInfo] Does any body have explanations about this type of exceptions?? Thanks for any suggestions :) I use the latest ecto (0.6.8) and ecto_ros (0.4.6) under ros indigo version. My computer run the ubuntu 1404 32bit system. Besides, SensorKinect is installed to make openni_launch work with Kinect. |
are you using openni or openni2 ? |
Where/how can I remap the topics being subscribed to by the object_recognition_server from the command line, if at all? Or do I have to create my own launch file and do so there? |
@pelferina When you run the server you give it a config file right? Doing: In that config file (https://github.com/wg-perception/tabletop/blob/master/conf/detection.table.ros.ork ) you specify the topics. I hope that helps. |
@vrabaud |
Thank you! That really helped a lot! |
Hi, I don't know if the problem is already fixed but I'm trying to follow the quickguide and the tutorials and I have some problems! Then, I have try to detect objects using tabletop and everything works well but when I have to the detect the object I have this error: [ INFO] [1457369669.064329642]: Initialized ROS. node_name: /object_recognition_server I use the latest ecto (0.6.11) and ecto_ros (0.4.8) under ros indigo distribution. |
@nosepas1 : For your first question: Are you sure that the output of your camera is working? Can you visualize its topics in ROS with rostopic echo <topic_name>? |
@hris2003 : The first issue it's true that I have some problems with the topics, many thanks. The second issue, I have add the mesh manually, I have capture nothing. And when I look in my db the mesh is there. Could it be something else? Thank you! |
@hris2003 : When I use freenect instead of openni all the topics are working well but the capture doesn't. Can I use freenect or it isn't compatible? |
@nosepas1 Which mesh did you add to the database? The coke mesh? Or did you prepare it yourself? Can I have a look at the mesh? |
@hris2003 I added the coke mesh from the tutorial. I get it from here: |
♫ This is the thread that never ends! Oh it goes on and on my friends! ♫ Error: I can't find the code/file (/build doesn't exist in my root directory) to see where the error is happening to try to fix it. I have read multiple forum posts about this error, but they all suggest edits to the code which I can't do as I can't find the source file ( Here is my gdb ouput. The line in question occurs after the subscribe to topics + thread exits.
According to Nosepas1's error output, the next line after the subscriber thread exits should read If this thread is better asked elsewhere, please tell me where. Thanks for your help! |
I fixed my error by changing the subscribed topic in the tod config file (see the above response from vraboud to pelferina) for the rgb_image_topic from Hopefully this helps anyone who runs into this problem! Now off to figure out why the program won't accept my uploaded coke can mesh during detection...... |
Hi~ I just got the same problem , when i'm run
Listening for key: s on imshow:save how could I solve this? I test this under my |
also get involved in the same problem, anyone solved it yet? |
In my case it worked after I |
I am flowing the quick start guide, but when I try detection_command, the program stops.
the error was
python: /usr/include/boost/smart_ptr/shared_ptr.hpp:418: T* boost::shared_ptr::operator->() const [with T = const sensor_msgs::Image_std::allocator]: Assertion `px != 0' failed.
https://groups.google.com/forum/#!topic/object-recognition-kitchen/mEW0Vn2NreM also reports the same problem.
I use openni_launch, but the object_recognition_core program always says the same error.
The text was updated successfully, but these errors were encountered: