Error in running object_recognition_core

[zaclizer]

Having installed from the ROS Groovy repositories, I run into this error when attempting to view the usage for object_recognition_core capture:

$ rosrun object_recognition_core detection -h Traceback (most recent call last): File "/opt/ros/groovy/lib/object_recognition_core/detection", line 3, in from object_recognition_core.utils.training_detection_args import common_create_parser, read_arguments_detector File "/opt/ros/groovy/lib/python2.7/dist-packages/object_recognition_core/utils/training_detection_args.py", line 5, in from object_recognition_core.boost.interface import ObjectDbParameters ImportError: No module named interface

Looking in $ /opt/ros/groovy/lib/python2.7/dist-packages/object_recognition_core/boost

the directory appears to be empty apart from an empty init file.

[vrabaud]

thx for reporting, trunk is fixed and 0.4.15 will be released soon

[vrabaud]

fyi, 0.4.15 got pushed to shadow-fixed

[Tatosx2]

Hello, I am running the following command: rosrun object_recognition_core object_add.py -n coke -d "A universal can of coke" and I am getting this error as a result (I've tried recreating the said files to no avail): Traceback (most recent call last): File "/home/tsquared/ork/src/object_recognition_core/apps/dbscripts/object_add.py", line 6, in from object_recognition_core.db import models File "/home/tsquared/ork/src/object_recognition_core/python/object_recognition_core/db/init.py", line 3, in from object_recognition_core.boost.interface import ObjectDbParameters, ObjectDbTypes, Documents, Models, Document ImportError: No module named interface Thanks in advance!

[awesomebytes]

Explaining a bit more of your setup may help understand your problem.

What Ubuntu version, what ROS version, why aren't you using the binary packages instead of a source code install.

On Mon, Apr 30, 2018, 16:03 Tatosx2 notifications@github.com wrote:

Hello, I am running the following command: rosrun object_recognition_core object_add.py -n coke -d "A universal can of coke" and I am getting this error as a result (I've tried recreating the said files to no avail): Traceback (most recent call last): File "/home/tsquared/ork/src/object_recognition_core/apps/dbscripts/object_add.py", line 6, in from object_recognition_core.db import models File "/home/tsquared/ork/src/object_recognition_core/python/object_recognition_core/db/ init.py", line 3, in from object_recognition_core.boost.interface import ObjectDbParameters, ObjectDbTypes, Documents, Models, Document ImportError: No module named interface Thanks in advance!

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/wg-perception/object_recognition_core/issues/7#issuecomment-385318866, or mute the thread https://github.com/notifications/unsubscribe-auth/ABpFdC35wkm3ySQ1-PudX8KzuerYvEh1ks5ttqkjgaJpZM4AX5us .

[Tatosx2]

Sorry my bad, ROS: Kinetic, Ubuntu 16.04. I temporarily "fixed" it by removing the indicated lines of code. Why should I be using the binary packages? I've now got a different error when I run: rosrun object_recognition_core detection -c  rospack find object_recognition_linemod/conf/detection.ros.ork I then get: Traceback (most recent call last): File "/home/tsquared/ork/src/object_recognition_core/apps/detection", line 6, in from object_recognition_core.pipelines.plasm import create_plasm File "/home/tsquared/ork/src/object_recognition_core/python/object_recognition_core/pipelines/plasm.py", line 4, in from object_recognition_core.io.voter import VoterBase File "/home/tsquared/ork/src/object_recognition_core/python/object_recognition_core/io/voter.py", line 5, in from object_recognition_core.ecto_cells.voter import Aggregator as AggregatorCpp ImportError: No module named voter

[awesomebytes]

Well, it very much looks like your setup is broken.

The binary packages... are tested and used by other people, so, they should work.

Give them a try:

sudo apt-get install ros-kinetic-object-recognition-*

And use a shell without your local workspace sourced to be sure you use those ones.

[Tatosx2]

Wow mate, thanks a lot, had no idea this existed,not sure as to how I was not aware. :) How would I go about installing this? So I just install these generically and make a new catkin work-space and go from there?

[awesomebytes]

How would I go about installing this?

sudo apt-get install ros-kinetic-object-recognition-*

So I just install these generically and make a new catkin work-space and go from there?

Yeah, open a new shell, where you only have source source /opt/ros/kinetic/setup.bash and just run your command again:

rosrun object_recognition_core object_add.py -n coke -d "A universal can of coke"