package
SLAM.Simulation
Landmark-level scenes without image processing
LandmarkCamera projects a fixed world scene into pixel/depth observations without image rendering or feature detection. landmarkGrid constructs a simple scene with stable row identities.
A real RGB-D reconstruction may seed a frozen scene after transformation into an explicit world frame. Preserve its calibration, initialization pose and feature identities. Reconstructed coordinates are not independent ground truth: keep simulated truth separate from uncertain estimator maps.
Members
| Name | Kind | Description |
|---|---|---|
| landmarkGrid | function | Fixed planar landmarks with stable row-major identities |
| observeLandmarks | function | Synthetic calibrated RGB-D measurements of a fixed scene |
| LandmarkCamera | block | Landmark-level RGB-D camera with explicit visibility and errors |
Modelica declarations and implementation
Show parameters, interfaces and source
Parameters, inputs, outputs, defaults and equations from the source file.
package Simulation "Landmark-level scenes without image processing"
annotation(Documentation(info = "<html>
<p><a href=\"modelica://SLAM.Simulation.LandmarkCamera\">LandmarkCamera</a>
projects a fixed world scene into pixel/depth observations without image
rendering or feature detection.
<a href=\"modelica://SLAM.Simulation.landmarkGrid\">landmarkGrid</a>
constructs a simple scene with stable row identities.</p>
<p>A real RGB-D reconstruction may seed a frozen scene after transformation
into an explicit world frame. Preserve its calibration, initialization pose
and feature identities. Reconstructed coordinates are not independent
ground truth: keep simulated truth separate from uncertain estimator maps.</p>
</html>"));
end Simulation;