Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Feature request : localization only mode. #150

Open
sandeepdutta opened this issue Feb 15, 2024 · 2 comments
Open

Feature request : localization only mode. #150

sandeepdutta opened this issue Feb 15, 2024 · 2 comments
Assignees
Labels
needs info Needs more information

Comments

@sandeepdutta
Copy link

Hello,

Would be great if the SLAM could be made to run in "localization" mode only.
That is , it will not publish any vo_pose or odom->base_link if it cannot localize .
If there is way to do this already then we will be happy to try.

Thank you
Sandeep

@swapnesh-wani-nvidia swapnesh-wani-nvidia self-assigned this Feb 21, 2024
@swapnesh-wani-nvidia swapnesh-wani-nvidia added the needs info Needs more information label Feb 21, 2024
@swapnesh-wani-nvidia
Copy link
Contributor

Could you clarify this? Are you looking to localize on a different map than the one created by Isaac ROS Visual SLAM?

That is , it will not publish any vo_pose or odom->base_link if it cannot localize .

@sandeepdutta
Copy link
Author

sandeepdutta commented Feb 21, 2024

Hello Swapnesh,

We are trying to localize in the map generated by VSLAM. We are using EKF to fuse wheel odometry & visual odometry, in a
premapped environment. If "Visual Tracking Lost" , the VSLAM algorithm reverts to "mapping" mode if it finds no loop closure and continues to publish visual_odometry with low "covariance" this confuses the EKF filter to rely on VSLAM more than it should. We are planning to update the "visual odometry" (set_odometry_pose) from the EKF fused odometry if visual odometry is lost.
If we had a "localize" only mode, then it would make the fusion much easier. If you have a better suggestion we will be
happy to try it.

Thank you
Sandeep

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
needs info Needs more information
Projects
None yet
Development

No branches or pull requests

2 participants