CloudCompare forum

Hello everyone,
I'm Flavia Sistilli, a PhD Student of Bologna University (Italy).
First of all, I would like to thanks all the developers for Cloud Compare.
I use CC for monitoring the accretions/erosion of 6 small areas of embryonic coastal dune (by photogrammetric survey).
I have some questions, but I have to advance that I'm not very expert in the field, so my questions might be banal.
Anyway, until now I have used the M2C tool to comparing the clouds,
I apply the mesh transformation (XY plane, with original normals) on the #Cloud 1 (the oldest one)
but I have noted that some time the response cloud is strange, like in the cloud in attachment:
there are some azure-blue patches that seem to be artifices, mistakes,
because they are on a quite homogeneous accreting zone and the forms too are strange.
Do you know this problem and why?

Meanwhile I have tried M3C2 plugin .
there are no patch in the response cloud, the result are ok but, this time, I have a graphic problem:
the response cloud show the correct difference but, visually, it's like #Cloud 1!
I would see the #Cloud 2 (the latest one) at the end of the process.
I have tried each possible combination of data, but I can't get what I want.
How can I do?
What could be the best setting/parameteres, for both M2C, M3C2 tools, for my situations of natural sand surface with vegetations?

Thanks in advance
Flavia Sistilli

Hi,

It seems I messed a little when validating your messages (I thought the second one was simply a duplicate of the first one). Can you post the images again?

Hi Flavia,

there are various ways to get your results on the 2nd cloud:
. Case 1: you select cloud#1 as the first survey, and cloud #2 as the second survey, in the M3C2 interface, and in the output tab you select "cloud #2". Note however that in doing so the point cloud produced by M3C2 will not be exactly the 2nd cloud, but a slightly smoothed version of it (the level of smoothing depending on the projection scale you selected).
. Case 2: if you want to absolutely preserve the positions of the 2nd cloud (i.e. no smoothing), you must invert the selection of clouds: cloud #1 will be the 2nd survey and cloud #2 the first survey, and in the output tab select "keep original positions". By default (and this is not changeable right now), if you impose a vertical difference, the direction is positive towards +Z. This means that you'll have the opposite sign compared to doing the first operation.

Hope this helps

Dimitri

Hi,
i'm sorry, excuse me,
i forgot to upload the image!

It's hard to understand exactly what's at stake, but those blue patches are generally due to wrong normals.

How does this mesh has been generated? Can you show it alone? (with normals activated)

Hi Daniel,
this is the mesh that I used. (Is this? It's correct?)
Usually I use the tool Edit->Mesh->XY plane -> NO update normal
(because when I import the clouds in CC, they are already "cleaned"and "repaired" with Geomagic Studio).
However, I also tried to update the normals and to use "best fitting plane" mesh
but the results are the same.
Maybe I should re-Compute the normals in CC, before the comparison?

Hi Dimitri,
thanks for your clarification.
therefore, there is not a solution for me.
maybe, the first solutions could be better to get a more clear reading of the image.

thanks so much for your time
Flavia

Could you post a snapshot (or even better: send me your file) of the mesh alone, and the cloud alone? If the gray (white/black) things I see are part of the mesh, then it looks clearly 'messy'.

hi Daniel,
i'm trying to uplod the mesh or the cloud, but evidently the files are too large!
How could I send them? with "We transfer"?or do you know another way?

Do you have a dropbox or google doc account? (you can then send me an email to cloudcompare [at] danielgm.nt)

Otherwise I can open an FTP account on CC's server (same thing, send me an email).

OK. I'm uploding the files on my dropdox.
i send you the link, as soon as the uplod is finished!
I'm sending you the .bin file with the 2 cloud + the mesh of the first one.

thanks
Flavia