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

Noise model for realsense d435 #58

Open
MarioCavero opened this issue Oct 28, 2024 · 0 comments
Open

Noise model for realsense d435 #58

MarioCavero opened this issue Oct 28, 2024 · 0 comments

Comments

@MarioCavero
Copy link

I have been using the rgbd information from the simulated d435 in ros2 humble. It wasn't until comparing a real realsense output that I adverted the quality of the depth map in simulation is almost ground truth (due to perfect conditions/no noise in simulation). And I'm only using it for 1m as max 1.5 meters.

The parameters of noise in the gazebo.xacro file seem to not do anything at all. I have already been thinking in processing ways as Multiplicative Gaussian Noise, in combination to edges and neighbor pixels. There is also a possibility to create a depth by projecting the diff of 2 simulated stereo, but it seems it is also very environment-dependent.
I have researched and apart from a Kinect adding noise repo, other users experiencing a similar issue with graphic engines in general that simulate with almost perfection. There exist some papers related to this but no documentation regarding this specific issue.

Any help would be appreciated.
image
image

And reality changes quite a bit in regards of the depth map, although the pointcloud seems more or less fine, I think it has too much detail in simulation.

Screenshot from 2024-10-27 16-08-52
Screenshot from 2024-10-27 14-30-50

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

No branches or pull requests

1 participant