Why is my meshed.ply model looks so distorted but fused.ply looks way better?

[Vinmean]

Hello,

I had to create a mesh file of a structure for my research. I walked around the structure to capture 178 pictures each of resolution 1280x720. The structure is a small rectangular building about 10 m length, 4 meters in breadth and about 4 meters in height. The fused.ply that comes out of the dense reconstruction process seems to be fine, but the meshed.ply which I want to use for my research doesn't seem to be textured properly. I would like to know what would be the likely cause for this to happen. I tried to follow the guidelines given in the COLMAP tutorials for capturing the pictures. There is also a number of 3D points outside of my region of interest. Can those points be removed and will that improve my mesh file?

 

Fused.ply

Meshed.ply

Thanks and Regards,

Vinmean

[Johannes Schönberger]

Hi,

If manual editing is an option, you could use Meshlab to remove the background structure and then run the dense_mesher on the edited file later. In general, I suggest to also perform proper texture mapping using Meshlab. I think, this is described here for COLMAP and Meshlab: https://www.youtube.com/watch?v=bDHJM6nAKtc

Let me know if you have other questions.

Best,

Johannes

[Vinmean]

Hi Johannes,

I tried the first option you gave me which is manually edit the sparse file to remove unwanted points and try performing the dense reconstruction. . After the sparse model is constructed along with camera positions, I exported the model as a Bundler  .out file. I deleted the points in Meshlab but I did not know how to import it back to continue with dense reconstruction. By manual editing, did you mean deleting the points inside COLMAP or using Meshlab?

[Johannes Schönberger]

No, what I meant is the following:

1. Run all steps, except the very last step ``dense_fuser`` that computes the meshed.ply from the fused.ply
2. Cleanup the fused.ply using Meshlab, save the file as fused-cleaned.ply
3. Run the last step ``dense_fuser`` with the fused-cleaned.ply as input and produce a meshed-cleaned.ply

Is that clear to you?

Cheers,
Johannes

> Fused.ply

>
>
>
>
>
> Meshed.ply
>
>
>
> Thanks and Regards,
>
> Vinmean
>
>

> --
> You received this message because you are subscribed to the Google Groups "COLMAP" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to colmap+un...@googlegroups.com.
> To post to this group, send email to col...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/colmap/d288f97a-80fa-4854-b2cf-65e52f4b11e7%40googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

[Vinmean]

Thank you Johannes, I am now able to get a better mesh. I ran the dense_mesher using the edited fused.ply file.

[Johannes Schönberger]

You might also want to adapt the parameters of the meshing, if you want to further fine-tune the results.

> On Mar 27, 2018, at 11:19 AM, Vinmean <iam...@gmail.com> wrote:
>
>
> Thank you Johannes, I am now able to get a better mesh. I ran the dense_mesher using the edited fused.ply file.
>

> --
> You received this message because you are subscribed to the Google Groups "COLMAP" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to colmap+un...@googlegroups.com.
> To post to this group, send email to col...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/colmap/8b1f64d1-d889-4d22-821e-0391bd2dcf5a%40googlegroups.com.