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

python sample code for aligning point clouds of an object captured from different angles #1941

Closed
bhomass opened this issue Jun 23, 2018 · 8 comments
Assignees

Comments

@bhomass
Copy link

bhomass commented Jun 23, 2018

Hi most of the sample code is in c++. I am working with python. Is there a python sample code available for aligning point clouds captured at different angles into one coherent point-cloud?

Associated question (I am a newby in point cloud work) - In general how many capture angles does one need to create a good quality 3D view?

@baptiste-mnh
Copy link
Contributor

Hi!
You can check the Open3D library which offer a lot of functionalities for 3D reconstruction.
The lib is very well documented, even if you're a beginner it should be easy to deal with it :)
http://www.open3d.org/docs/tutorial/Basic/icp_registration.html

You should have a least ~75% of overlapping between two captures to get a good reconstruction.
The more you points of view, the best the reconstruction is.

@dorodnic
Copy link
Contributor

opencv_contrib/rgbd also has some useful algorithms for point-cloud alignment.

@bhomass
Copy link
Author

bhomass commented Jun 25, 2018

@dorodnic would that be kinecfusion.? All the code is in c++ in that lib. Is there a simple rule for wrapping ?

@dorodnic
Copy link
Contributor

I personally tried odometry_evaluation that accepts two RGBD images and calculates the transformation between them (using either ICP or other methods).
I used C++ but you can most likely get the same functionality from python. Try pip install opencv-contrib-python, usually OpenCV APIs are translated very closely between C++ and python.

@bhomass
Copy link
Author

bhomass commented Jun 26, 2018

Thank you all for the answers.

@baptiste-mnh is there a ready api in open3d which can detect that the overlap is less than 75% and therefore suboptimal?

@RealSense-Customer-Engineering
Copy link
Collaborator

[Realsense Customer Engineering Team Comment]
Hi @bhomass,

Are you able to have the implementation according to @baptiste-mnh's and @dorodnic's suggestions?

@bhomass
Copy link
Author

bhomass commented Jul 20, 2018

Another team member will move forward with this. Will get back to this forum if there are questions.

@RealSense-Customer-Engineering
Copy link
Collaborator

[Realsense Customer Engineering Team Comment]
Hi @bhomass,

I will close this first. If any further question, you still can re-open or create another ticket.

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

No branches or pull requests

4 participants