SLAM on Parkour Car

Many years ago, I tried running a sparse SLAM algorithm clled RSLAM on a Jetson TK1 computer which was at the time the main computer for my Parkour Car platform.

Eventhough, I was able to run RSLAM successfully but because of low update rate (4-5 estimates/sec) I decide to move on with Vicon feedback instead. One can always optimize available sparse slam methods (like frame resizing, number of features, … ) to get a higher update rate but since that was not focus of my research I didn’t work on it further.