Class SemanticSegmentationLabeler

Labeler which generates a semantic segmentation image each frame. Each object is rendered to the semantic segmentation image using the color associated with it based on the given SemanticSegmentationLabelConfig. Semantic segmentation images are saved to the dataset in PNG format.

Only one SemanticSegmentationLabeler can render at once across all cameras.

Inheritance

object

CameraLabeler

SemanticSegmentationLabeler

Implements

IOverlayPanelProvider

Inherited Members

CameraLabeler.enabled

CameraLabeler.hudPanel

CameraLabeler.overlayPanel

object.ToString()

object.Equals(object)

object.Equals(object, object)

object.ReferenceEquals(object, object)

object.GetHashCode()

object.GetType()

Namespace: UnityEngine.Perception.GroundTruth

Assembly: solution.dll

Syntax

[Serializable]
public sealed class SemanticSegmentationLabeler : CameraLabeler, IOverlayPanelProvider

Constructors

Name	Description
SemanticSegmentationLabeler()	Creates a new SemanticSegmentationLabeler. Be sure to assign labelConfig before adding to a PerceptionCamera.
SemanticSegmentationLabeler(SemanticSegmentationLabelConfig, RenderTexture)	Creates a new SemanticSegmentationLabeler with the given SemanticSegmentationLabelConfig.

Fields

Name	Description
annotationId	The id to associate with semantic segmentation annotations in the dataset.
labelConfig	The SemanticSegmentationLabelConfig which maps labels to pixel values.

Properties

Name	Description
description	A human-readable description of the labeler
label	Interface that should be defined by a class that wants to be able to provide a image to the overlay panel.
overlayImage	Interface that should be defined by a class that wants to be able to provide a image to the overlay panel.
supportsVisualization	Labelers should set this in their setup to define if they support realtime visualization of their data.
targetTexture	The RenderTexture on which semantic segmentation images are drawn. Will be resized on startup to match the camera resolution.

Methods

Name	Description
Cleanup()	Called when the Labeler is about to be destroyed or removed from the PerceptionCamera. Use this to clean up to state.
OnBeginRendering()	Called just before the camera renders each frame the the labeler is enabled and ShouldCaptureThisFrame is true.
Setup()	Called just before the first call to OnUpdate() or OnBeginRendering(). Implement this to initialize state.

Events

Name	Description
imageReadback	Event which is called each frame a semantic segmentation image is read back from the GPU.