Page 1 of 1

Report of comparison between 2 point cloud or point cloud/mesh

Posted: Wed Apr 15, 2015 8:36 am
by Paul L.
Hi,

I am new on CloudCompare.
I want to compare two point cloud or point cloud/mesh and I would like to obtain a report with all measures differences, not only standard deviation and maximum/minimum.

Is it possible?

Thanks,

Paul.

Re: Report of comparison between 2 point cloud or point cloud/mesh

Posted: Wed Apr 15, 2015 9:18 am
by daniel
All measures? As clouds generally have millions of points/values, this would be a bit huge ;).

Well, you can still export the cloud (and associated distances values) as an ASCII file that can be easily imported in Excel for instance - if there are not too many points.

You can also spawn labels on various points (hold SHIFT + click on a point) so as to get a global overview.

Re: Report of comparison between 2 point cloud or point cloud/mesh

Posted: Wed Apr 15, 2015 2:36 pm
by Paul L.
Yes all measures. The point cloud has only 150 points.

A comparison cloud is created? Because when I compare the two clouds, I only have the window with "distance mean, average, ...", and no file is created when I close it.

Re: Report of comparison between 2 point cloud or point cloud/mesh

Posted: Wed Apr 15, 2015 3:41 pm
by daniel
Have you clicked on the 'compute' button?

See http://www.cloudcompare.org/doc/wiki/in ... d_Distance

If yes, the compared cloud will get a new scalar field (with the per-point distances). Then you can close the tool dialog, select this cloud and display the color scale, export it, etc.

Re: Report of comparison between 2 point cloud or point cloud/mesh

Posted: Wed Apr 15, 2015 8:27 pm
by Paul L.
Thanks !

When I export the compared point cloud, I have this report format (ascii) :

100110.34839630 5006589.53300095 123.66700006 255 0 0 -0.004026 0.076671 0.025557 0.996729

How read lines?
X Y Z R G B ... and after there are 4 elements, What numbers represent differences?

And how CloudCompare calculate statistics? Only in Z ?

Thanks for your answers !

Re: Report of comparison between 2 point cloud or point cloud/mesh

Posted: Thu Apr 16, 2015 7:13 am
by daniel
I guess your file already had some scalar fields at loading time? Hard to tell what the values are though.

My first guess would be that the last one is the distance. To be sure you can add a header line to your ASCII file (check the ASCII output dialog).

For base statistics, we use the (3D) distances. If you want to separate the distances on X, Y and Z, check the 'split X,Y and Z components' option in the C2C distances computation dialog.

Re: Report of comparison between 2 point cloud or point cloud/mesh

Posted: Thu Apr 16, 2015 12:48 pm
by Paul L.
Ok I add a header line in the ASCII file.

A last question : When i want compare cloud/mesh the 'split X,Y and Z components' option is not accessible (in grey)? Why?

Re: Report of comparison between 2 point cloud or point cloud/mesh

Posted: Thu Apr 16, 2015 1:44 pm
by daniel
Well, it's mainly due to the fact that the algorithms are not the same.

You may also have noticed that the C2C distances are always positive while the C2M distances are signed by default.

Re: Report of comparison between 2 point cloud or point cloud/mesh

Posted: Fri Apr 17, 2015 3:06 pm
by Paul L.
It is possible to compute differences in Z (split X Y and Z) between mesh (reference) and cloud?

Re: Report of comparison between 2 point cloud or point cloud/mesh

Posted: Fri Apr 17, 2015 4:07 pm
by daniel
Nope :D