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

Is there any way to remove depth data of target's side? #8306

Closed
hailsong opened this issue Feb 4, 2021 · 10 comments
Closed

Is there any way to remove depth data of target's side? #8306

hailsong opened this issue Feb 4, 2021 · 10 comments
Labels

Comments

@hailsong
Copy link

hailsong commented Feb 4, 2021

  • Before opening a new issue, we wanted to provide you with some useful suggestions (Click "Preview" above for a better view):

  • All users are welcomed to report bugs, ask questions, suggest or request enhancements and generally feel free to open new issue, even if they haven't followed any of the suggestions above :)


Required Info
Camera Model { R200 / F200 / SR300 / ZR300 / D400 }
Firmware Version (Open RealSense Viewer --> Click info)
Operating System & Version {Win (8.1/10) / Linux (Ubuntu 14/16/17) / MacOS
Kernel Version (Linux Only) (e.g. 4.14.13)
Platform PC/Raspberry Pi/ NVIDIA Jetson / etc..
SDK Version { legacy / 2.. }
Language {C/C#/labview/nodejs/opencv/pcl/python/unity }
Segment {Robot/Smartphone/VR/AR/others }

Issue Description

<Describe your issue / question / feature request / etc..>

I'm developing example for reconstruct 3d pointcloud of target and I used ICP (point to plane). I worked well but I found some pointcloud of target's side make the output bad. It's maybe the background noise of realsense and I tried to apply removing outlier tool of Open3D but it didn't work.
Here's photo of my output
image

My question is 'Is there any tool to remove pointcloud of target's side?' In 2D image, I used morphology operation for erode... is there any tool like this?

@MartyG-RealSense
Copy link
Collaborator

MartyG-RealSense commented Feb 4, 2021

Hi @hailsong I considered your image carefully. My understading of outlier filtering is that it is suited for removal of detached floating clumps of point cloud data that are separated from the main image. Therefore it may not help if the points on an edge are untidy but firmly connected to the main body of the cloud.

Could you describe the method that you are usin to generate the Open3D cloud with RealSense please?

@hailsong
Copy link
Author

hailsong commented Feb 9, 2021

@MartyG-RealSense Hi! Thanks for responding.
Firstly, I got pointclouds through pipeline and cropped my ROI (in this case, the target was a person). I saved it in .ply file.
image
(pointcloud from realsense)
image
(cropped pointcloud data, I saved it to .ply file)
Than I opened .ply file through open3d.io.read_point_cloud and apply ICP Multiway registration.
And I found some noise of pointcloud and I guessed this is noise from background (or maybe motion blur?)
image
gray pointclouds are the trash pointcloud (it's on my nose) (and its color is from background. my labs wall is gray color.)

@hailsong
Copy link
Author

hailsong commented Feb 9, 2021

@MartyG-RealSense
I tried realsense's built-in filter RSM - threshold remove (?) to reduce this kind of noise.
image

I can reduce it but I lose some correct pointcloud data..

@MartyG-RealSense
Copy link
Collaborator

If you are creating the ply file in the RealSense Viewer and then opening it in Open3D, you could try setting the Depth Units option in the Viewer to 0.0001 instead of the default 0.001. This will create a more zoomed-in view but can reduce depth noise significantly.

You can find the Depth Units option under the Controls sub-section of the Stereo Module options

In the case in the link below, I used a 0.0001 depth unit scale to create a textured point cloud in the Viewer's 3D mode that had much less noise than when using the default 0.001 scale.

#8228

@hailsong
Copy link
Author

image
Your solution makes it better! but I still have some noise of it.

@MartyG-RealSense
Copy link
Collaborator

MartyG-RealSense commented Feb 10, 2021

Under the Depth Control section of the Advanced Controls options in the Viewer, does setting DS Second Peak Threshold to zero reduce the amount of loose dots on your image?

image

@MartyG-RealSense
Copy link
Collaborator

Hi @hailsong Do you require further assistance with this case, please? Thanks!

@hailsong
Copy link
Author

@MartyG-RealSense thanks a lot!

@MartyG-RealSense
Copy link
Collaborator

Is your problem solved now, please? Thanks!

@MartyG-RealSense
Copy link
Collaborator

Case closed due to no further comments received.

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

No branches or pull requests

2 participants