This week I did extensive research into optical localization algorithms to consider as an alternative to wireless localization methods. I learned about SIFT, SURF, ORB, and Optical Flow as approaches to tracking the position and orientation of a camera in space from its video feed. I was able to put together such a working demo, placing 3 imaginary points in space that stayed in place in the world as the camera moved. Ultimately, we decided to continue on the path of the Wi-Fi and Ultrawideband-based methods, with a possible use of optical methods as a fallback or for sensor fusion.
My progress is currently ahead of schedule, as I will be working on the hardware of the user’s tag device, the major component of which turned out to be in the department inventory. So that saved a lot of time against ordering it online and having to wait for it.
In the coming week, I plan to setup the hardware for the user tag and learn how to program and interface with the board. This way, when it comes time to implement the signal-based localization algorithms, I will be familiar with both the software and hardware.