Segmentation Camera

The segmentation camera provides ground truth labels for a scene.

Implementation details

This camera generates a label value for each object in the scene, instead of an intensity/color image. More specifically, it will assign every Renderer in the scene a unique color in the output.

This isn't used to model a real camera, but as a ground truth. For example, you can use this to train a machine learning algorithm that segments a scene into objects.

Inputs

Input	Description
ResolutionH	Horizontal number of photocells on the photosensor array.
ResolutionV	Vertical number of photocells on the photosensor array.
SensorSizeH	Physical width of the photosensor array, in millimeters.
SensorSizeV	Physical height of the photosensor array, in millimeters.
FramesPerSecond	Number of captured frames per second.

Outputs

Output	Description
Frame	A render texture resident on the GPU storing object labels.
OutTranscode	Provides a context for executing sampling custom passes, which then passes through the GPU processing pipeline.
FieldOfView	The field of view, driven by focal length and sensor size.