Visual Inertial Odometry (VIO) information for graduate student
-
Hello all,
I am working with three Seeker drones as part of a graduate thesis and I'm looking for some background on the VIO implementation running on the Seeker drones. My understanding is that there is some Open CV based processing happening on the VOXL computer, and that VIO messages are published via ROS for PX4 to interpret and report as the local position estimate. However, what is this VIO algorithm exactly comprised of? Is it an open source implementation that I can investigate and play with myself, or is this homegrown code specifically for the Seeker drones?
My end goal lies in developing a SLAM algorithm, so I'd love to hear whatever knowledge you all have about how visual data is processed.
Any insight or links for further reading are much appreciated.
Thank you!
Nick B
