Drone exterior + laser scanner interior + key area photo combine together texture issue

Community Community & Industry Discussion
1

Hi there,

I try to model a room from both indoor and outdoor. the exterior data captured by drone and the interior data captured by the laser scanner, also the interior key area I user camera to take photos in order to present details. the data can be aligned and meshed perfectly, only after texture I found the interior texture is terrible, seems the photo taken by drone project to laser scanner color, how can I keep indoor laser scanner color?

BTW, I can’t disable laser scanner’s color during texture since I only taken partial indoor photos.

Thanks

2

Hi @JohnathanZhang1

Try to disable the external images for texturing and then texture the interior. If it will be the same, then the external images don’t affect the laser scan texturing. The laser scans are not ideal for such work and it is possible it won’t be nice.

What kind of laser scanner have you used?

3

Hi @OndrejTrhan

Thanks for your quick response. I use Faro P200 laser scanner. When I try to combine photo and laser scan together, seems there is light cast from outdoor, all windows and doors frame color project to the wall and ground, make the indoor texture dirty. Also I found if I just import laser scan data, after color mesh model and export to .las point cloud, compare to the original laser scan color data is not good.

Is there any special parameters I set to wrong？

Thansk

4

Hi Johnathan,
unfortunately I am not aware of such setting. The texturing using laser scans is the general issue and the best option here is to have properly captured images also from the interior for texturing.

5

Hi Otrhan,

Is there any workflow to mesh/texture exterior drone and interior laser component separately, then merge them together. I think this way the color won’t interfere each other.

Thanks

6

Not in RealityScan, but it should be doable in other 3rd party application for merging meshes/point clouds.

To get the models in the same scale and position I would use the data in RealityScan and then use the sources for meshing separately to get two models (one from images and one from laser scans).

7

So you will convert LS point cloud to mesh model with color? but the generated mesh model is too big, what’s the best setting to simplify the LS mesh model?

Thanks