Skip to content

Modified tool of the TUM RGB-D dataset that automatically computes the optimal scale factor that aligns trajectory and groundtruth. Useful to evaluate monocular VO/SLAM.

Notifications You must be signed in to change notification settings

raulmur/evaluate_ate_scale

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 

Repository files navigation

evaluate_ate_scale

Script to compute the Absolute Trajectory RMSE error for the TUM RGB-D dataset, that automatically aligns the scale of the estimated trajectory with ground-truth. It is useful to evaluate trajectories generated by SLAM or VO algorithms. See the original tool and how to use it here: http://vision.in.tum.de/data/datasets/rgbd-dataset/tools#evaluation

Trajectory and ground-truth are aligned computing a Similarity Transformation (Rotation, Translation and Scale) with the method of Horn: B. K. P. Horn, “Closed-form solution of absolute orientation using unit quaternions,” Journal of the Optical Society of America A , vol. 4, no. 4, pp. 629–642, 1987. pdf

About

Modified tool of the TUM RGB-D dataset that automatically computes the optimal scale factor that aligns trajectory and groundtruth. Useful to evaluate monocular VO/SLAM.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages