matlabbe

Well, I've checked in find_object code and PCL was not used, so I removed the ROS dependency in this [commit](https://github.com/introlab/find-object/commit/ae7b79629b95c42dfd74653e43fd86f6cecf575c) so that PCL is not required to build find_object_2d. Thx...

Yes, but in this example we added some fake TF just to show relation between object and the map (in case we are using a mobile robot). Remove the following...

Copy it in your home directory, then modify it. You can launch any launch files directly: ``` $ roslaunch find_object_3d.launch ```

Is there an object detected in find object ui?

Can you show the RGB and depth images? Looks like there are no valid depth values on the object detected.

To use the depth, find-object should be subscribed to it. The find_object_3d.launch shows how to do that. Make sure the depth image is actually published. For example, removing not needed...

See the warning: > maybe object is too near of the camera or bad depth image Use rqt_image_view to show the depth image, and take a screenshot showing the find-object...

The rqt_image_view is not required, but it helps to debug when you get warning about no depth values detected on the objects. Not sure why TF names are not shown...

You should change the `map_frame_id` value, not the name: ```xml ``` The transform you shown, we can see that the object is at 78 cm in front of the camera....

Like the [documentation](http://wiki.ros.org/find_object_2d) is describing: > Objects detected formatted as [objectId1, objectWidth, objectHeight, h11, h12, h13, h21, h22, h23, h31, h32, h33, objectId2...] where h## is a 3x3 homography matrix...