0
$\begingroup$

Can you please point me to a data set that can be used to validate monocular visual SLAM?

My vision for what this data set looks like is a series of frames, with ground truth pose, ground truth pixel measurements of features, and ground truth positions of all features. I imagine that this would be a synthetic dataset generated in a simulator.

If you think this exists a link would be appreciated, I couldn't turn anything up myself.

Cheers and thanks!

NOTE I am not looking for something like kitti or the tum data sets or some RGB-D datasets, I want ground truth poses, feature measurements, and feature positions.

EDIT I only need this to validate the optimization - not all the other parts of the classic VSLAM pipeline fyi

$\endgroup$
0
$\begingroup$

I have a github project that extracts the visual inertial simulator from OpenVins to its own mini project.

https://github.com/Edwinem/VI-Simulator

I think it fulfills your needs. Gives you access to the ground truth poses, features,feature positions, ... .

If you ignore the inertial data portion then it is also not too hard to create your own. Just generate a random pointcloud and use some of the multi view geometry equations to project them to a random camera pose.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.