feat: Add virtual detection area

[jruebsam]

Reference to a related issue in the repository

733

Add a description

see Issue

Take this checklist as orientation for yourself, if this PR is ready for the Change Control Board:

• My suggestion follows the style and contributors guidelines.
• I have taken care about the message documentation and the fields and enums documentation.
• I have done the DCO signoff.
• My changes generate no errors when passing CI tests.
• I have successfully implemented and tested my fix/feature locally.
• Appropriate reviewer(s) are assigned.

[thempen]

Other Models & Harmonization Group:

• The Polygon3D should not be restricted to 2 dimensions. The restriction should be made in the VirtualDetectionArea documentation.
• The position to the virtual sensor coordinate system should be clarified.

[PhRosenberger]

In general, I support the idea of the virtual detection area.

However, some more discussion is needed. Let's discuss the open points and questions I raised, shall we?

[PhRosenberger]

Why did you choose Vector3d / cartesian coordinates for the Polygon3d? - I suggest to use Spherical3d instead, because imho it is more common to use angles and ranges to describe the field of view of a sensor.

[PhRosenberger]

Actually, I wonder why to use Polygon3d anyways, as you could describe the FoV with repeated Spherical3d coordinates already. Did you insert this intermediate step with the surfaces for a specific reason?

In other words: Why do we limit the Virtual detection area to be a composition of surfaces rather then a more flexibel solution? - Sensors like the Lovox Tele-15 definetly need more flexibility to be described.

[thomassedlmayer]

I think, Vector3d was used because that's what 3d engines offer for setting up such polygons. Converting this to spherical coordinates for transmission is just another step, I guess. So for our use case of visualizing such area Vector3d is completely fine.

The use case for this PR is just a quite basic visualization of such sensor cones, so we did not want it to become more complicated than it has to be. But I would agree, that the intention of this detection area should be mentioned in the description more specifically.

[jruebsam]

I would say in general the Vector3d approach is more generic.
Using spherical3d defines something which always contains the origin at 0.
So for defining some abstract surface in 3d it will not be possible to use Spherical3d.

The question is how generic should it be ?

On the other side as Thomas mentioned the Vector3d Format is much better suited towards
rendering applications.

[jruebsam]

edit: I just realized you can also exclude 0 from spherical ...

[pmai]

The prosposed wording changes might make the intent clearer.

[pmai]

CCB 2024-03-25: Merge with suggested changes.

[jdsika]

@pmai
We have the base_polygon. We should have defined this at the time?

//
// \brief Polygon in 2 dimensions
//
// A polygon in 2 dimensions which contains a list of vertices.
//
message Polygon2d
{
    // A list of vertices
    //
    repeated Vector2d vertex = 1;
}

Reviewed for v3.7.0