This section of the webpage is designated for information about my senior design project.  My senior design project is based on the growing use of action cameras to film sports and extreme activities.  Many people now attach small cameras to their helmet or other sports equipment in order to take video from amazing angles never seen before.  My goal for this project was to take the hassle of using a camera and trying to mount it somewhere completely out of the equation.  To do this my vision is to create an autonomous drone platform that will be able to carry the camera for you and film you from amazing angles.  This quadcopter will be able to “lock on” to a person and then follow them at a specified height as they do an activity.  For more information on this project and to see the progress being made please feel free to go to my “Squadcopter” homepage where details of the project will be posted.

Quadcopter Details

In the implementation of this project there are three main parts; the quadcopter platform, the action camera, and the autonomous system.  The easiest to accomplish of these three parts is the action camera.  The action camera that we chose to use on our quadcopter is the GoPro Hero 3 Silver edition.  This is a compact and lightweight camera that is capable of recording high definition videos (1080p) for up to two full hours.  The GoPro is a natural choice for this project, however many other action cameras would have done the job.  The camera is mounted on the top of the quad copter so that it has an unobstructed view of the space in front of it.  Also, the camera is mounted on a vibration damping platform that reduces distortion created by the high frequency vibration of the motors.

The quadcopter platform was a little more difficult to complete than the simple camera set up.  We started with a frame made of G10, which is essentially a fiber glass composite that is very tough and lightweight.  This frame is made by Lumineer and it is called the QAV250 (the 250 designation means that the frame is 250mm diagonally).  There are also four Lumineer brush-less motors that we selected to give the quadcopter enough lift to carry all of the equipment.  Those four motors are connected to four Lumineer electronic speed controllers or ESCs that handle the precise control of the motors.  The ESCs are then connected to the main flight controller, which we selected to be the Naze32.  The Naze32 is perfect for our project because of its compact size and its ability to be stabilized using PID controllers, which allows us to tune the quadcopter for different payloads. We connected the ESCs and flight controller to the battery through the use of a power distribution board or PDB, which allows us to effectively power all of the components using one lithium ion battery.

The autonomous system is the most difficult part of the project to complete.  This system is made up of a brain and a sight system.  The brain of the system is the arduino mega, which has many PWM output that will be useful in the future.  The sight system is a camera called a PixyCam, which is capable of detecting blocks of color.  The first step in this system was to place the arduino between the radio receiver and the flight controller.  This will initially act as a bus that mocks and passes along the receivers output to the flight controller.  The next step is to the modify the output that the arduino is sending to the flight controller, so that we can control the quadcopter independently of the radio controller.  After this is accomplished we can start using the PixyCam as the eyes of the system.  From this we can know how far the quadcopter is from a color coded object that we want to follow.  The camera outputs x and y coordinates, height and width of the colored object, and rotation of that object.  After processing this data with the arduino, we will then begin to convert this data into commands for the quadcopter.  By doing this we can wean the quadcopter off of radio controls and start focusing on flying only based on what the PixyCam sees.  The final goal is for this platform to be fully autonomous and to follow a color coded object flawlessly.