CloudCompare forum

Hello,

I am a student doing research on debris flows and have collected LiDAR data of a mountain side using a Teledyne Optech Polaris LiDAR. My goal is to remove vegetation from the mountain side in order to eventually calculate volumes of sediment, deposited or accumulated, before and after a storm. I have tried using many different software's to remove vegetation (e.g Atlascan, QT Modeler, etc.),but I have high hopes for CloudCompare. I was attempting to work with the CANUPO plugin until I stumbled across the new 3DMASC plugin. My question is would using the 3DMASC plugin be able to complete my goal of removing vegetation and leaving a 'bare earth model' of the mountain side?

Also if 3DMASC is the right way to go about this, where can I download the plugin and if someone can share the steps on how to set it up would be great!

I am running CloudCompare version 2.12.1. Thanks.

Hi,

3DMasc is part of the latest 2.13.beta version.

To remove vegetation, the options are, in order of complexity: CSF, Canupo and 3DMasc (see the plugins page: https://www.cloudcompare.org/doc/wiki/index.php/Plugins).

Sadly I am not an expert so I won't be able to help you much further. However, there are dedicated threads on this forum for the Canupo and 3DMasc plugins. Or you might try to contact the authors directly.

Thanks for your response. Hopefully someone knowledgeable can chime in.

So I have loaded the vegetsemi.prm using the Canupo Classify Cloud tool and have generate the below screenshot. I did not create my own classifiers. Not sure what the next step is after this. How can I remove the vegetation from this point? Also why is there a blue "blind" spot on the point cloud?

To segment the points you'll have to use the 'Edit > Scalar fields >Filter by value' method. And using a default classifier only makes sense if you have the same units as what was used to train it... Maybe it would be worth creating your own classifier?

Thanks for a quick response. I was able to use the Filter by value method successfully. Now, once the vegetation has been stripped, is there a way to create a bare earth model?

Well, you can use the 2.5D Delaunay triangulation, or the PoissonRecon plugin (which is more tricky to use as you will need first to compute clean normals, but more robust to noise).

Thank you.

I used the scissor tool to classify vegetation and ground, but I got an error saying that I need to create meshes to use CANUPO classification? What am I doing wrong here.

Can you paste the text or a screenshot of the error you get? Because Canupo definitely works (and requires) clouds.