Bird's-Eye Scope

Visualize sensor coverages, detections, and tracks

Description

The Bird's-Eye Scope visualizes aspects of a driving scenario found in your Simulink® model. Using the scope, you can:

  • Inspect the coverage areas of radar and vision sensors.

  • Analyze the sensor detections of actors, road boundaries, and lane boundaries.

  • Analyze the tracking results of moving actors within the scenario.

To get started, open the scope and click Find Signals. The scope updates the block diagram, finds signals representing aspects of the driving scenario, organizes the signals into groups, and displays the signals. You can then analyze the signals as you simulate, organize the signals into new groups, and modify the graphical display of the signals.

For more details about using the scope, see Visualize Sensor Data and Tracks in Bird's-Eye Scope.

Open the Bird's-Eye Scope

From the Simulink model toolbar, click the Bird's-Eye Scope button . If instead you see a button for a different model visualization tool, such as the Simulation Data Inspector or Logic Analyzer , click the arrow next to the displayed button and select Bird's-Eye Scope.

Your most recent choice for data visualization is saved across Simulink sessions.

Parameters

expand all

Global Settings

To access the global settings of the Bird's-Eye Scope, from the scope toolstrip, click Settings.

Longitudinal axis limits, specified as a [min, max] vector.

Tunable: Yes

Lateral axis limits, specified as a [min, max] vector.

Tunable: Yes

Selection matrix used to extract the positions of tracked objects, specified as a 2-by-n matrix of zeros and ones. n is the size of the state vector for each tracked object in the scenario. The scope multiplies the selection matrix by the state vector of a tracked object to return the (x, y) position of the object.

  • The first row of the matrix corresponds to the x-coordinate stored within the state vector.

  • The second row of the matrix corresponds to the y-coordinate stored within the state vector.

This parameter applies to signals from a Multi Object Tracker block that were initialized by a linear Kalman filter. The state vector format depends on the motion model used to initialize the Kalman filter. For more details on these motion models, see trackingKF and Linear Kalman Filters.

The default selection matrix is for a 3-D constant velocity motion model. In this motion model, the state vectors of tracked objects are of the form [x;vx;y;vy;z;vz], where:

  • x is the x-coordinate of a tracked object.

  • vx is the velocity of a tracked object in the x-direction.

  • y is the y-coordinate of a tracked object.

  • vy is the velocity of a tracked object in the y-direction.

  • z is the z-coordinate of a tracked object.

  • vz is the velocity of a tracked object in the z-direction.

Multiplying the state vector by this selection matrix returns only the first element of the state vector, x, and the third element of the state vector, y.

[1,0,0,0,0,0; 0,0,1,0,0,0] * [x;vx;y;vy;z;vz] = [x;y]

Tunable: No

Selection matrix used to extract the velocities of tracked objects, specified as a 2-by-n matrix of zeros and ones. n is the size of the state vector for each tracked object in the scenario. The scope multiplies the selection matrix by the state vector of a tracked object to return the velocity of the object in the (x, y) direction.

  • The first row of the matrix corresponds to the x-direction velocity stored within the state vector.

  • The second row of the matrix corresponds to the y-direction velocity stored within the state vector.

This parameter applies to signals from a Multi Object Tracker block that were initialized by a linear Kalman filter. The state vector format depends on the motion model used to initialize the Kalman filter. For more details on these motion models, see trackingKF and Linear Kalman Filters.

The default selection matrix is for a 3-D constant velocity motion model. In this motion model, the state vectors of tracked objects are of the form [x;vx;y;vy;z;vz], where:

  • x is the x-coordinate of a tracked object.

  • vx is the velocity of a tracked object in the x-direction.

  • y is the y-coordinate of a tracked object.

  • vy is the velocity of a tracked object in the y-direction.

  • z is the z-coordinate of a tracked object.

  • vz is the velocity of a tracked object in the z-direction.

Multiplying the state vector by this selection matrix returns only the second element of the state vector, vx, and the fourth element of the state vector, vy.

[0,1,0,0,0,0; 0,0,0,1,0,0] * [x;vx;y;vy;z;vz] = [vx;vy]

Tunable: No

  • Select this parameter to display short signal names (signals without path information).

  • Clear this parameter to display long signal names (signals with path information).

Consider the signal VisionDetection within subsystem Sensor Simulation. When you select this parameter, the short name, VisionDetection, is displayed. When you clear this parameter, the long name, Sensor Simulation/VisionDetection, is displayed.

Tunable: Yes

Signal Properties

These properties are a subset of the available signal properties. To view all the properties of a signal, first select that signal from the left pane. Then, from the scope toolstrip, click Properties.

Transparency of the coverage area, specified as a real scalar in the range [0, 1]. A value of 0 makes the coverage area fully transparent. A value of 1 makes the coverage area fully opaque.

This property is available only for signals in the Sensor Coverage group.

Tunable: Yes

Scale factor for the magnitude length of the velocity vectors, specified as a real scalar in the range [0, 20]. The scope renders the magnitude vector value as M × Velocity Scaling, where M is the magnitude of the velocity.

This property is available only for signals in the Detections or Tracks groups.

Tunable: Yes

Limitations

  • Referenced models are not supported. To visualize signals that are within referenced models, move the output of these signals to the top-level model.

  • Rapid accelerator mode is not supported.

  • If you initialize your model in fast restart, then after the first time you simulate, the Find Signals button is disabled. To enable Find Signals again, from the model toolstrip, click the Disable Fast Restart button .

  • The Bird's-Eye Scope does not support visualization in a model that contains:

    • More than one Scenario Reader block.

    • A Scenario Reader block within a nonvirtual subsystem, such as an atomic or enabled subsystem.

    • A Scenario Reader block that is configured to output actors and lane boundaries in world coordinates (Coordinate system of outputs parameter set to World Coordinates).

More About

expand all

Tips

  • Make sure that all sensor blocks have unique sensor identifiers. These identifiers are specified in the Unique identifier of sensor parameter of Vision Detection Generator and Radar Detection Generator blocks. If two sensor blocks have the same identifier, the Bird's-Eye Scope visualizes sensor data from only the first sensor block specified in the model. Duplicate sensor identifiers can occur when you add a default Vision Detection Generator block and Radar Detection Generator blocks to your model. The default Unique identifier of sensor parameter value for these sensor blocks is always 1.

  • To find the source of a signal within the model, in the left pane of the scope, right-click a signal and select Highlight in Model.

  • You can show or hide signals while simulating. For example, to hide a sensor coverage, first select it from the left pane. Then, from the Properties tab, clear the Show Sensor Coverage check box.

  • When you reopen the scope after saving and closing a model, the scope canvas is initially blank. Click Find Signals to find the signals again. The signals have the same properties from when you last saved the model.

  • If the simulation runs too quickly, you can slow it down by using simulation pacing. From the Simulink model toolbar, select Simulation > Pacing Options. Then, select the Enable pacing to slow down simulation check box and decrease the simulation time to less than the default of one second per wall clock second.

Introduced in R2018b