matlabbe

Yes, you could modify [tf_example_node.cpp](https://github.com/introlab/find-object/blob/master/src/ros/tf_example_node.cpp#L76-L79) to publish markers using the `pose` position in `/map` frame. Find-Object publishes objects detected in `/camera` frame. TF is used to transform that position in...

You could copy [tf_example_node.cpp](https://github.com/introlab/find-object/blob/master/src/ros/tf_example_node.cpp) in your package. The ros dependencies of your package would be `tf` and `find_object_2d`. cheers

You don't really need Qt to make it work, you can remove the header and change that [line](https://github.com/introlab/find-object/blob/84cf6280ac9aadca3145b524453b710bd34cea5e/src/ros/tf_example_node.cpp#L58) to not use Qt.

Make sure you did `source /opt/ros/kinetic/setup.bash` and `source ~/catkin_ws/devel/setup.bash`. I don't have a kinetic machine to test with, but the following is working with a Kinetic docker image: ``` $...

Can you do `catkin_make -j1 VERBOSE=1` to have more info? Do you have more than 1 opencv installed?

Possible solution: https://github.com/introlab/find-object/issues/32

We use cv::VideoCapture class from OpenCV, the image query is there: https://github.com/introlab/find-object/blob/ef0c565b7e3b3a3e78ae5424a7536cf3dd3e9101/src/Camera.cpp#L126 There are couple solutions in this post: https://stackoverflow.com/questions/30032063/opencv-videocapture-lag-due-to-the-capture-buffer

Humm, make sure you are on ros2 branch of rtabmap_ros: https://github.com/introlab/rtabmap_ros/blob/ros2/package.xml (find_object_2d is not a dependency). Find-Object is indeed not released on ROS2. If there is an interest, I could...

Hi Manu, The features are contrast dependent. They are computed in grayscale version of the image. Inverting black in white in a image would result in different features. Well, look...

Hi Manu, In the video they are using SIFT + Color descriptors. You may have similar results with SIFT alone. I don't know how KAZE performs in comparison of SIFT....