Method Question

I understand that it's intended that we use vision-based navigation to control the drone.

However, I was wondering if it's plausible that we try using a pre-constructed map of important target points within range of possible flight paths. I propose this because I thought color recognition (as presented by the team from last year) might be too little information or too easily mistaken for surrounding objects of the same color.

I was thinking that each point would be assigned a 3D location in relation to an origin (the origin could be the center of the music room), and that the drone could use its internal accelerometer, sonar distance sensor, and altimeter to keep track of its own location.

A flight path could be created by sending a list of target points in order of occurrence along the path, which the drone could then follow.

Do you think this could be done, or are there too many possibilities for error in the drone's capability to keep track of its own location?

Research Task 1 (06/22/15 - 07/03/15)

Linux

1. Start learning Linux on your own by watching YouTube videos [18 videos] or following a free course "Linux Foundation" from EdX. Since Linux is an extremely popular Operating System (OS) used in the academic research environment. You might want to pick it up as soon as possible. We plan to run our programs of drone vision and drone navigation on a Linux platform, so, getting familiar with it is essential for our STEM project.
2. You will need to download and install Ubuntu 12.04 (LTS) on your computer, or use a Linux laptop provided by school. 

AR Drone 2.0

1. AR Drone Developer Guide, Parrot. [chap. 1-3, or 17 pages] This guide will give you an overview of the structure, operations, hardware, software, and communication of AR Drone 2.0.
2. Download the FreeFlight app from the App Store (for iPhone/iPad) or from the Play Store (for Android). Watch AR Drone video Tutorials #1 and #2, and learn to fly the drone! You will need to use an AR Drone from school for your flight.

Computer Vision & OpenCV

1. A.R Drone Vision-Guided Searching (2013), Derek Long. [chap 1-2, 18 pages] The first two chapters of this dissertation gives you a brief introduction to the important concepts in computer vision and OpenCV. We are going to use OpenCV to process the video acquired by the drone, and perform "vision-based navigation".

Please take electronic notes while you are studying the materials, watching the videos, or browsing through the web. Each team will present their learning later in the summer meeting.