Hmm, probably this question is to special or bad described?

I only want to get the best possible result during my calculations for the vertices, faces and normals.

But maybe it is not only a thing on my side but also on the way how wikitude is doing the transformation from the opengl coordinate system to a ?real world coordinate system, probably wgs84 or pseudo mercator? or vice versa?

I only want to get the best possible result during my calculations for the vertices, faces and normals.

But maybe it is not only a thing on my side but also on the way how wikitude is doing the transformation from the opengl coordinate system to a ?real world coordinate system, probably wgs84 or pseudo mercator? or vice versa?

Hi Robert,

Sorry for the late response. I'm not sure if you already read the parts of our documentation that I linked below. Especially the first one might be interesting for you:

We will add more information about this topic in our documentation very soon.

Best regards,

Andreas Schacherbauer

Hi Andreas,

thx so far, I'm excited to see the new documentation.

So in general, I have to set for the DBS values like minScalingDistance: 0.1 and maxScalingDistance: 300 to disable this function more ore less. Because I don't want to scale up my pipes for readability. I think the more difficult part is to find a correct value for the SDU.

The documenation has an example for a drawable but not for a 3D-Model.

For drawables attached to GeoObjects, 1 SDU is defined as the screen-size of an object which is exactly 1 meter high and located 10 meters away from the viewer. If a drawable's width is defined as 5 SDU, it will exactly overlap an object which is 5 meters wide and is located in 10 meters distance.

For the 1 SDU example the documentation say 1 meter high in 10 meter, but the 5 SDU example say 5 meter wide in 10 meter. So width or length?

How is this related to a 3D-Model? Do I have to transform every vertices to SDU values?

Greetings

Robert

© 2021 Wikitude, a Qualcomm company

## Robert Gregat

Hi, I have a couple of questions regarding the creation of 3D-Models.

1.) What kind of coordinate system is used in wikitude? For the positioning it is necessary to use WGS84 (geography) Coordinates. On the other side it is also possible to translate an object but with metric values, which makes sense because to do that with geographic values is horrible with all these decimal places.

2.) With the first question/assumption in mind I would think that I need a coordinate reference system which is valid for the most part of the world, for example the pseudo mercator reference system, to create my 3D-Models. I know, that every 3D-Model has an independent coordinate system which is anchored around one point (the origin) which will be used later to position it in the ar scene. But to create the 3D-Models geographical correct I need the correct coordinate system to do my calculations.

3.) Why I think of the pseudo mercator reference system? Well it is a metric coordinate system and I guess that you transform the geography coordinate for the positioning (where a 3D-Model is anchored) in your ?AR metric reference system? some how.

4.) But what is still a question mark in my mind, is the rotation and actual the scale which I have to respect if I want a true representation of my 3D-Model in the real world. What I mean with this statement. If I have a simple rectangle in the real world with the dimension of 3*2 m and a certain rotation, well than I want to create a 3D-Model which fits at the end exactly on this real world rectangle (I purposely left out the accuracy of the positioning of a mobile device).

I know this are a few questions but I hope that you can shed some light.