I am working on trying to automate the landing sequence of a multicopter to help it land autonomously on a charging station. My thought is that the copter would takeoff, fly its waypoints, return and loiter over the launch area. From that point, a vision based system would help align the copter and keep it on track as it descends to the charging cradle.
GPS does not have the resolution needed to hit an exact spot for me. The live video feed can be transmitted down to the ground station computer using a wifi IP camera as it will be in close range. Using SimpleCV to image process and sending MAVlink protocals through a telemtry link and python scripts I can have the ground station computer do the heavy computing and image processing and send navigation commands back to the copter.
Does anyone know of any existing work on this or would want to contribute?
Replies
Have you made progress on this? I'm interested.