University of Twente Student Theses

Login

Drift correction using a multi-rate extended Kalman filter

Knuppe, R. (2018) Drift correction using a multi-rate extended Kalman filter.

[img] PDF
2MB
Abstract:This work proposes an algorithm to improve an existing positioning system in a virtual reality application. This existing system uses the inertial based XME (Xsense MVN engine) which is prone to drift. Drift is a real problem, because accidents can happen when the virtual world is misaligned with the designated play area in the real world. After considering several approaches we settled on using a second positioning system in the form of an optical inside-out approach. In order to deal with the opportunistic and desynchronised nature of this method we developed a multi-rate extended Kalman filter for the sensor fusion. The filter was designed with two update steps, one for each positioning system. One update step takes the pose of the OPS and the other a delta pose of the current and previous XME pose. With this approach a test was done where the designed OPS error model and the delta approach were compared to simple alternatives. The test used a third, proven, pose estimation system (the optitrack) to provide ground truth poses. The results show that the OPS error model doesn't improve much over a static error. The delta approach is very preferable compared to the non delta approaches tested. The results also show that the algorithm does correct the drift.
Item Type:Essay (Master)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:54 computer science
Programme:Interaction Technology MSc (60030)
Link to this item:http://purl.utwente.nl/essays/74650
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page