RTabMap and Find-Object with two RGB-D cameras

[Hubble]

Hello everyone!

I reproduced a demo_find_object example on my project. I attached my *.launch file and ROS graph rosgraph.svg. Tell me please how can I achieve the same result as in "Robot mapping with Find-Object" tutorial? My work space is: Ubuntu 16.04 LTS, ROS Kinetic 1.12.7 and RTabMap 0.11.13. Also, can I run one find-object on two cameras and change the detector in the launch file?

[Hubble]

I ran rviz instead rtapmapviz and there the object is displayed. The rostopic list shows two topics, is it possible to subscribe rtabmapviz on them? And can I adjust the pixels size, as in rviz?

[matlabbe]

Hi,

find_object_2d assumes one image topic, so you may create two find_object_2d nodes in different namespaces.

It is not possible to display object detections in rtabmapviz. Note that with RVIZ, you can get the main information shown in rtabmapviz. For the "pixels" size in rtabmapviz, see Preferences->3D rendering->Point size. When paused, you can also press keys "-" or "+" to increase/decrease point size.

cheers,
Mathieu

[Hubble]

Thanks for your answer!