So I am a little confused and just want to make sure that I have this down correctly, as I have been in the research stage for awhile now looking for options but need to move forward.
The goal is to have an autopilot control the quad, go a short distance and use some object recognition software to identify some object placed amongst several other objects.
I have the EB3 and quad already assembled. Now, can this be accomplished using an autopilot such as px4, some RPI object recognition script and Qgroundstation? Alternatively I thought it may be possible to use px4 and ROS, can ROS be used as a ground station sort of like mission planner or QGS, and also control the quad to do things on top of what the autopilot is already performing?
I guess I am getting confused with where the actual coding takes place, within ROS, or on the RPI image flashed to the EB3. I have read some things about using MAVLINK with QGC also. How exactly will the quad know what to do after it has arrived at its destination?
Sorry for the long winded post and any help is greatly appreciated.