Edit and exporte the point cloud (neglecting fusion step)

Hello, I have a Revopint Mini and I would like to export the point cloud to .ply file in a straightforward way before the fusion process. Edit it is also impossible before this step and it would be interesting for me …
Could you help me with this?
Moreover, I have noticed that the minimum step for the fusion step is higher for the 5.1.1 RevoScan version in comparison with the previous version which is also a step-back in this line.
In any case, if I simply edit a and export point cloud data without the fusion process it would be very good for my application,
could you help me?

Hi!
To export your raw point cloud as .ply for using in other programs like cloudcompare you have to export it as .ply by clicking here

For being able to fuse at 0.02mm with mini you have to activate “standard” mode first (under “fusion”)! advanced mode doesn´t let you to choose this lower point distance.

Thank you Ivan! I did not consider that the standard mode had more options than the advanced one.
Now, in relation with my main question, I want to export (and edit if possible) the cloud points before the “Fusion” task, however the “Export” option is not avaliable if the “Fusion” taks it is not done :frowning:
For my application the Fusion task introduces noise so I want to export the cloud points before …
The cloud data is there but I cannot export it …

1 Like

you are welcome!
I see…no, there isn´t such option. :slightly_frowning_face:
But, what makes you think fusion process introduces noise?
My experience with mini: noise level depends on different factors:

  1. is mini calibrated properly?
  2. Scanning environment → the darker, the better
  3. Surface of object → Color/ Glossiness/ reflectivity
    → for best results using scanning spray with even and as thin layer as possible, like Attblime AB2 with 3 microns achievable applying with airbrush

Scan with Mini, question from newbie - #8 by PopUpTheVolume

You can’t export unfused frame cells since it is not editable in any software on a planet .
To have an editable point cloud you need to fuse the frame cells first and there is no other way around . That’s how the process goes .

If you want to have less noises please use the advanced mode for fusing or increase the pitch point level when using standard mode , most of scanned objects don’t need 0.02mm fusion pitch point as that is half the hair wide size , and even visible with an naked eye .

To avoid noises do not scan reflective or shiny materials , prepare the surface to avoid noises . That includes most plastics and metals that produces micro reflections .

If you scan small objects , and don’t use marker mode , switch the side of the scanning plate or don’t use it at all .

1 Like

Thank very much for your response.
I cannot believe that unfused frame cells are not editable because they are a point cloud … they are list of XYZ point positions (at least), I only want to process by myself.
The 0.02 mm process is the factor which is creating some kind of artifacts, the unfused point cloud is more accuracte in my particular case (from a visual point of view because I cannot process this data).
If I can process by myself I can average or process noise by myself too, in my case i am scanning a very simple surface from upside and I do not neet move it, however I want to research if I can get better denoising with my own algorithms.
About …
“If you scan small objects , and don’t use marker mode , switch the side of the scanning plate or don’t use it at all .”
I was also thinking about this, my object is in fact small and I was consideing what is the best method in a multiple frame acquisition (I am not moving the objetc or the scanner) to avoid additional artifacts … I am using the “marker mode” because I think that my object have not any clear detail to be stable in a multiple frame acquisition. It seems plausible for me … I need to research about that.
However, to research I need access to this undused data
Thank you again to both of you

1 Like

instead of usimg marker mode you might also try feature mode using crumpled paper under your object as explained and described here.

then you can compare the results to scanning using marker mode.please share your findings.cheers!

1 Like

better to apply for the SDK to get the hardware unlock software and then capture your points directly with revopoint 3D viewer app that is on Revopoint’s github

1 Like

Crumpled paper, good trick. Thank you Ivan

2 Likes

this is probably the best option to me. I will apply, thank you, I did not know this option!

1 Like

@fatxutxa

Awsome, keep us posted on your findings with the SDK :slight_smile:

I have applied but haven’t had time to mess with it yet, don’t know if you are aware of my app the Revoscan Flayer (Frame Player) that allows you to inspect the RGB images of the frame capture. I have been trying to find out how the depth map (dph files) are put together but so far was only able to visualize the map with rawpixels.net

Check my post and if you have anything to contribute i’ll be very gratefull :slight_smile:

@fatxutxa

See if the frame2pointcloud helper code for my Revoscan Flayer is of any help to you

1 Like