List view
We need a computer vision algorithm that can recognize the Russian text on the blue background of the mast. Most likely, we will not know the orientation of the mast, so Flight will be implementing an orbiting algorithm for moving around the mast. This means that the drone will be moving in a circle while attempting to identify the text, most likely with the camera pointing toward the mast at each point. This task does not involve writing the orbiting algorithm, but it may be useful information for considering noise.
Overdue by 5 year(s)•Due by November 21, 2020•3/3 issues closedWe need to have a way for Flight to retrieve and interpret the data from Vision. This needs to be scalable such that Vision can add/change data and Flight can easily get that data. The objective by the deadline is to have telemetry describing what the drone sees (obstacles, objectives, other relevant data). This will allow us to verify how well our algorithms work during autonomous flight at a Flight Test. Vision's task is to continue working on the existing Pipeline class to ensure it is compatible with a live camera feed. Vision and Flight must coordinate to ensure that the Pipeline class can be used in the state machine to output relevant telemetry during different stages of the competition (e.g. searching for the mast is only relevant after finishing the initial pylon laps).
Overdue by 5 year(s)•Due by November 14, 2020•1/1 issues closedThe drone needs to be able to identify obstacles in its path, and filter out noise and faraway objects. There should be a pipeline to easily interpret what obstacles the drone sees, as well as a pipeline to send relevant data to flight.
Overdue by 5 year(s)•Due by January 30, 2021•3/5 issues closedThis project involves reacting to obstacles in the environment appropriately and safely. - Detect obstacles using computer vision. - Pipeline for sending data between vision and flight. - Drone responds to new obstacles in flight path.
Overdue by 5 year(s)•Due by January 30, 2021Once the drone arrives at the mast, we use computer vision to identify the location and orientation of the module. This involves not just identifying a module, but ensuring we can locate that module when on the mast, accounting for noise or other variables. This task does not involve interacting with or replacing the module.
Overdue by 5 year(s)•Due by January 30, 2021•14/17 issues closedOnce the drone finishes the pylon laps, it needs to navigate to the mast to prepare to replace the module. This task does not involve removing the module. - Use provided GPS coordinate to fly to the mast - Employ orbiting algorithm to fly around the mast (this will prepare the drone to search for the mast text) - Create config file to organize our GPS coordinates, rather than defining constants in each file
Overdue by 5 year(s)•Due by November 21, 2020•2/2 issues closedThe drone should be able to complete all laps, both to the mast and returning to the takeoff zone.
Overdue by 5 year(s)•Due by November 21, 2020•0/1 issues closed