Combine two rectified point clouds

B970418

January 09, 2023 06:49

Hello, I have two D435i. I have corrected them through the checkerboard, remapped the left and right images, and completed the homography to stitch the two corrected real-time videos. What I have to do now is to process the two The two point clouds of the real-time video are synthesized into the same point cloud, what should I do?

Comments

7 comments

MartyG

January 09, 2023 11:09
Hello, it sounds as though you have two separate point clouds but have not yet combined them into a single cloud and want to know how to do it. Is this correct, please?
0

Comment actions Permalink
B970418

January 09, 2023 11:33
Yes, I have two rectified separate point clouds and I would like to know how to merge them
0

Comment actions Permalink
MartyG

January 09, 2023 11:37
Which programming system are you using with your point clouds please (C++, Python, PCL, Open3D, etc).
0

Comment actions Permalink
B970418

January 09, 2023 11:43
My pre-work is all done using python and opencv
0

Comment actions Permalink
B970418

January 09, 2023 11:48
If you don’t have python-related methods, you can also provide methods in other programming languages, I can try to research, thank you very much
0

Comment actions Permalink
MartyG

January 09, 2023 12:05
It is possible to combine point clouds in Python using the RealSense SDK's rs2_transform_point_to_point instruction, which can combine data from different viewpoints using a procedure called an affine transform. There are not many references for using this instruction though. The link below may be helpful.

https://github.com/IntelRealSense/librealsense/issues/5583#issuecomment-570805578

In regard to other platforms, Intel have a tutorial for combining point clouds from two RealSense cameras using ROS.

https://github.com/IntelRealSense/realsense-ros/wiki/Showcase-of-using-2-cameras

Intel also have a guide for stitching two pointclouds using C++ and MATLAB.

https://github.com/IntelRealSense/librealsense/tree/master/wrappers/pointcloud/pointcloud-stitching

Views from multiple RealSense cameras can be combined with the commercial software tool RecFusion Pro.

https://www.recfusion.net/index.php/en/features

Two point clouds could be combined in the Open3D platform using a procedure called ICP.

https://github.com/IntelRealSense/librealsense/issues/10477

PCL has a method called Normal Distributions Transform.

https://pointclouds.org/documentation/tutorials/normal_distributions_transform.html
0

Comment actions Permalink
B970418

January 09, 2023 12:08
Thank you very much for your help.
0

Comment actions Permalink

Please sign in to leave a comment.

Comments

Didn't find what you were looking for?