Filtering a 3D point cloud

I have attached a 3D point cloud here . You can call the function using:

   DisplayReconstructedScene( RimPointCloud_filt_lst_sq, J1 );

It looks like this:

How can I process this further and remove all noise? I want only the circular object's surface to be detected. Kindly let me know if further information is required. I want to remove those tiny groups of data points which are away from the main data points.

I have already tried pcdenoise. It didn't show any difference to my results.

[Jemma Greve]

did you find a solution?

[Meghana Dinesh]

Yea, pcdenoise did not help me, even after trying various parameters. I had to write a custom code. However, it would be easy if I knew this .

[DS]

Hi, have you solved your problem? If so, could you give some hints on what you have done? Thank you very much in advance!

Answer 1

[abuzer]

Maybe too late but someone need for answer.

Latest Matlab has Point cloud processing toolbox.

http://www.mathworks.com/help/vision/point-cloud-upsampling-downsampling-and-denoising.html

pcdenoise function will suit for your work. But you need to change parameters for neighbor and sigma threshold.

ptCloudOut = pcdenoise(ptCloudIn)

http://www.mathworks.com/help/vision/ref/pcdenoise.html

cheers.