CloudCompare forum

Okay I understand. Sadly to do this in CC you would first need to 'calibrate' your photo (i.e. determine the exact position of the camera relatively to the scanner). Then you would be able to apply the camera 'viewport' and trace the polygons directly in CC (with the scissors tool).

The only missing step is how to calibrate your camera... I know that Trimble Realworks for instance is able to do this (you pick equivalent points in both the picture and the cloud and once you have enough pairs the software computes the best parameters). You can even export the result to VRML and then import it in CC (it's a bit tricky so if by any chance you get there you'll have to ask me again for directions ;).

It would be very cool to be able to do this directly in CC but I don't know how difficult it really is and how long it would take me...

Hi Daniel,

Thanks for your answer. I appreciate your help. I have a new question, very important.

I would like to know how to calculate the numerical difference (i.e., what was the amount of change that took place in terms of volume loss meters cubed) when I compare two point clouds?

When I subtract one point cloud from the other, I have a new image (point cloud) that shows spatially, where 'change' took place. But, I want to know how much ice was lost (in terms of volume) from the glacier.

In total I have around seven point clouds. They are all scanned at different time periods. I want to look at the evolution of the glacier's front section by comparing two point clouds...and I would really like to quantify the volume change from the glacier. I just don't know which options to select in cloud compare.

Hope you can help! Merci, thank you again so much!

Hi again Daniel,

When I draw a polygon on to the point cloud, it preserves the x,y,z coordinates of the 'bounding box' - but this does not give me the true volume of the object inside the bounding box. The ice bergs are irregularly shaped, but the bounding box is a cuboid. Is there any way to extract the polygons and determine the volume inside the polygon, rather than the bounding box?

Computing a volume without a mesh is very hard.

You have two options:
- the hardest but the most accurate: create a closed mesh by merging the front of the disappearing volume (subset of the 'before' cloud) and the back (subset of the 'after' cloud). And then use PoissonRecon to build the 3D mesh, and eventually use 'Edit > Mesh > Compute Volume'. You'll have to compute clean normals on the clouds, with normals pointing in opposite directions for both clouds!
- the easiest: transform your cloud(s) so as to get the normal to the change parallel to X, Y or Z and then use the new 2.5D Volume Computation tool (http://www.cloudcompare.org/doc/wiki/in ... .5D_Volume).

Hi Daniel,

Thank you once again for all your answers, and your patience. New question,

Q. Is it possible to draw a polygon onto the point cloud and save the polygon coordinates as a txt file?

I know it's possible to do this with point labels, but does it work in the same way for polygons?

In one of my previous messages, I mentioned the "bounding box". When I draw a polygon, CC puts a "bounding box" over the polygon and reports the x,y,z coordinates of the box. I want to extract the x,y,z coordinates of the polygon instead, so I can report the volume inside the polygon.

Any ideas?

(I'm using 2.6.2 CC now)

If you want to draw a polygon that uses some of the cloud points as vertices, then you can use the 'Point List Picking' tool (there's an option to export the selection as a polyline).

Otherwise you can use the trick of the 'Scissors tool' (select the cloud, start the Scissors tool, then draw the 2D polyline and export it).

In both cases you'll be able to export the vertices coordinates by selecting the polyline you've created and saving it as a 'SALOME Hydro Polyline' (it's a very simple file format where only the vertices are saved as clear text).