package

SLAM

Visual inertial localization, mapping and loop closure

Visual localization, mapping and loop closure. Image acquisition, matching, registration and calibration are in Vision.

Localization: visual observations and registration uncertainty.
Mapping: anchored landmarks and spatial indexing.
LoopClosure: keyframes, retrieval and verification.
PoseGraph: measured factors and graph optimization.
Fusion: visual aiding for the navigation ESKF.
Simulation: fixed scenes and synthetic RGB-D observations.
Examples: application and sensor examples.

Inertial is a world-additive ES15/Schmidt reference. It uses different tangent and injection conventions from the navigation ESKF; they require an explicit adapter to exchange internal states. Full graph execution must be checked against the selected compiler.

Members

Name	Kind	Description
AdvanceFastSLAM	function
AdvanceFastSLAMIntervals	function
D435FastSLAMInitialize	model
D435FastSLAMIntervals	model
D435FastSLAMStep	model
Inertial	package	Inertial propagation and Schmidt relative-pose correction
Examples	package
Fusion	package	Visual measurements for the CogniPilot navigation estimator
InitializeFastSLAM	function
Localization	package	Calibrated relative RGB-D localization and navigation interfaces
LoopClosure	package	Keyframe selection, retrieval and geometric loop verification
Mapping	package	Landmark projection, identity and persistent map storage
PoseGraph	package	Pose-graph optimization, gauge handling and estimator commits
RGBDFastSLAMInitialize	model
RGBDFastSLAMInterface	model
RGBDFastSLAMIntervals	model
RGBDFastSLAMReset	model
RGBDFastSLAMStep	model
Simulation	package	Landmark-level scenes without image processing

Modelica declarations and implementation

Show parameters, interfaces and source

Parameters, inputs, outputs, defaults and equations from the source file.

package SLAM "Visual inertial localization, mapping and loop closure"
  annotation(uses(Vision, Estimation, LinearAlgebra, RigidBody, Control),
    Documentation(info = "<html>
    <p>Visual localization, mapping and loop closure. Image acquisition,
    matching, registration and calibration are in
    <a href=\"modelica://Vision\">Vision</a>.</p>
    <ul>
    <li><a href=\"modelica://SLAM.Localization\">Localization</a>: visual observations and registration uncertainty.</li>
    <li><a href=\"modelica://SLAM.Mapping\">Mapping</a>: anchored landmarks and spatial indexing.</li>
    <li><a href=\"modelica://SLAM.LoopClosure\">LoopClosure</a>: keyframes, retrieval and verification.</li>
    <li><a href=\"modelica://SLAM.PoseGraph\">PoseGraph</a>: measured factors and graph optimization.</li>
    <li><a href=\"modelica://SLAM.Fusion\">Fusion</a>: visual aiding for the navigation ESKF.</li>
    <li><a href=\"modelica://SLAM.Simulation\">Simulation</a>: fixed scenes and synthetic RGB-D observations.</li>
    <li><a href=\"modelica://SLAM.Examples\">Examples</a>: application and sensor examples.</li>
    </ul>
    <p><a href=\"modelica://SLAM.Inertial\">Inertial</a> is a world-additive
    ES15/Schmidt reference. It uses different tangent and injection conventions
    from <a href=\"modelica://Estimation.StrapdownINS.ESKF\">the navigation
    ESKF</a>; they require an explicit adapter to exchange internal states.
    Full graph execution must be checked against the selected compiler.</p>
    </html>"));
end SLAM;