Distance calculation

[younell]

Hello dear Mates,
I'm really happy to discover this very useful software and I thank you for your outstanding work.
I'm using cloudcompare in order to calculate the deformation field by comparing 2 cloud points ( I've too many cloud points for deformed surfaces and I'm not sure if I can speficify the developpement of this deformation with comparing all this points cloud at the same time, anyway). So I proceed as follows:
1-Loading 2 points clouds.
2-Aligning these clouds.
3-Calculate the distance between them using Cloud to cloud distance ( which gives me an absolute distance (I'm wondering if there's a way to calculate the real distance & how can I precess to do so)
4- after calculating this distance I'm wondering if there's a way to Export the distance between these 2 points clouds ( for each coordinate )
for example ( coordinate of my cloud point-X dis-Y dis-Z-dist )

Waiting forward to hearing from you.
Regards.

[daniel]

Indeed by default cloud-to-cloud distances are always 'unsigned' as we have no information on the 'side' of the surface (i.e. where is the matter).

To get a signed distance you can either compute a mesh on the reference cloud (if it's easy to do of course). Otherwise you should give a look at the qM3C2 plugin (http://www.cloudcompare.org/doc/wiki/in ... PluginM3C2).

And for your last question (item 4): be sure to check the 'split X, Y and Z components' checkbox in the distance computation dialog. Then select the compared point cloud only and export it as an ASCII file (File > Save as). In an ASCII file you'll get per line:

Code: Select all

X, Y, Z, absolute distance, dist X, dist Y, dist Z

[younell]

Thank you Daniel for your detailed answer, but I'm using Cloudcopmare V 2.5.3 but I can' find qmc32 plugin, and I'm wondering if there's a way to include it to my version?
Looking forward to hearing from you.
Best regards.
Youcef

[daniel]

Can't you use the 2.5.4 version?

You can try to copy the qM3C2 plugin DLL from the 2.5.4 version into the 'plugins' directory of your version. But you'll fall into the gray zone and anything can happen ;)

[younell]

Thanks Daniel for your help.
I just calculated distances between cloud points , and after saving the scalar field as an ASCII file I noticed after reloading it that I have a kind of average of my previous Scalar field.
So according to you what was the reason behind that and how what the best format to make sure that I will note lose some data ( if there's a solution to get my data back I'd be definitely glad to hear it)

Best regards.
Youcef

[daniel]

To avoid any loss of accuracy when saving in ASCII format, just use a high precision (12 for instance). Of course it will make bigger files.

To get the same result with less space, the PLY format is your friend (but you won't be able to open it easily in other softwares).