Detailed Description

Renderer for simulating RGB-D sensor output from 3D models.

This class provides functionality to render synthetic RGB-D frames from a set of 3D meshes and a camera configuration. It supports semantic labeling and rendering of surface coordinates.

Usage:

Set the camera using setCamera().
Set the 3D model(s) using setModel().
Configure label mode and surface coordinate rendering as needed.
Call render() with a pose to generate a SimulatorFrame.

Important aspects:

Supports face and vertex label modes for semantic segmentation.
Can render surface coordinates (xyz maps) in addition to color, depth, and label images.
Not thread-safe.

Example:

SimulatorRenderer renderer;
renderer.setCamera(cameraComponent);
renderer.setModel(meshes);
renderer.setLabelMode(SimulatorRenderer::FaceLabel);
renderer.setRenderSurfaceCoordinates(true);
auto frame = renderer.render(poseMatrix);
auto colorImg = frame->colorImage();
auto depthImg = frame->depthImage();

Public Types
enum	LabelMode { FaceLabel = 0 , VertexLabel = 1 }
	Source of labels to use for creating the label maps. More...

Public Member Functions
	SimulatorRenderer ()
	Constructs a SimulatorRenderer.

	~SimulatorRenderer ()
	Destructor.

void	setCamera (RGBDDataComponent *camera)
	Sets the camera configuration.

RGBDDataComponent *	camera () const
	Returns the current camera configuration.

void	setModel (const std::vector< Mesh * > &meshes)
	Sets the 3D model(s) to be rendered.

std::vector< Mesh * >	model () const
	Returns the current 3D model(s).

void	setLabelMode (LabelMode mode)
	Sets the source of label for creating the semantic segmentation image.

LabelMode	labelMode () const
	Returns the source of label for creating the semantic segmentation image.

void	setRenderSurfaceCoordinates (bool value)
	Enables or disables rendering of surface coordinates (xyz maps).

bool	renderSurfaceCoordinates () const
	Returns whether rendering of surface coordinates is enabled.

std::unique_ptr< SimulatorFrame >	render (const mat4 &T)
	Renders a frame from the current model and camera at the given pose.

Member Enumeration Documentation

◆ LabelMode

enum LabelMode

Source of labels to use for creating the label maps.

Enumerator
FaceLabel	Uses the labels of faces.
VertexLabel	Uses the labels of vertices.

Member Function Documentation

◆ setLabelMode()

void setLabelMode ( LabelMode mode )

inline

Sets the source of label for creating the semantic segmentation image.

The mode is not updated if the provided value is incompatible.

◆ render()

std::unique_ptr< SimulatorFrame > render ( const mat4 & T )

Renders a frame from the current model and camera at the given pose.

Returns: Unique pointer to the rendered SimulatorFrame.

The documentation for this class was generated from the following file:

RGBD/Include/ImFusion/RGBD/SimulatorRenderer.h