First, create catkin workspace and include everything in this repository in the src folder
Next, download the model folder from team's Google drive ([email protected]). Copy the model folder and paste it inside the cv_tracker folder
Replace the /flea3/launch/stereo_node.launch with the stereo_node.launch from team's Google drive ([email protected]).
Include the two yaml files from team's Google drive ([email protected]) in your ROS_HOME
*To launch the image capturing module:
roslaunch flea3 stereo_node.launch
*To launch the detection/classification module:
roslaunch cv_tracker ssd.launch
-To view detection information on person objects:
rostopic echo /obj_person/image_obj
-To view detection information on car objects:
rostopic echo /obj_car/image_obj
*To launch the depth measure module:
roslaunch depth_measure depth_measure.launch
NOTE:
Please ONLY push the working code back to Github
Depth information published by depth_measure (or elas_ros) does not seem correct...
I think the "depth measure" module can be used for sensor fusion task. So no need to have another seperate node for sensor fusion?
Let's plan to use Rviz for visualization for now