Skip to content

Implementation of different Kalman Filters to estimate the state of the quadrotor using optical flow and IMU.

Notifications You must be signed in to change notification settings

jagennath-hari/Kalman-Fusion-Quadrotor-State-Estimation

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

32 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Kalman Fusion: Quadrotor State Estimation

To estimate the state of a quadrotor using different sensor a roboust state estimation technique is required. The gold-standard Kalman Filter was used to fuse data from sensors such as Intertial Measurement Unit(IMU), Vicon(Motion tracking cameras) and Visual Odometry(Optical Flow). The entire project has been divided into $3$ parts(a, b, c). Part (a) implements the Extended Kalman Filter(EKF) to fuse data from IMU and Vicon. Part (b) shows the effectiveness of RANSAC during visual odometry to remove outliers. Finally, part (c) shows the effectiveness of Unscented Kalman Filter(UKF) at the cost of computational speed by fusing the data from optical flow and IMU.

Kanade–Lucas–Tomasi(KLT) feature tracker using ORB

Experiment and Inference

The $3$ submodules contain each of the parts(a, b, c) described above.

Click HERE for Part (a).

Click HERE for Part (b).

Click HERE for Part (c).

Conclusion

The EKF proves to be the most efficient tool for fusing different sensors. The added benefit of high precision and fast computational speed makes it the most optimal sensor fusion technique for mobile robots lacking high-end on board computer. UKF on the other hand improves the accuracy of state estimation using sigma point propogation by sacrificing computational speed. The UKF is a viable state estimation technique for those robots having sufficient computational power, moreover visual odometry and IMU tends to drift over time, the usage of the UKF can improve the accuracy of state estimation when robots are traversing over longer durations.

About

Implementation of different Kalman Filters to estimate the state of the quadrotor using optical flow and IMU.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published