robokudo.descriptors.camera_configs.components

Classes

DepthComponent	Camera component with configuration for depth input data.
ColorComponent	Camera component with configuration for color input data.
TfComponent	Camera component with configuration for transform handling.
StableViewpointComponent	Camera component with configuration for viewpoint stabilization.
WorldDescriptorComponent	Camera component with configuration for world descriptors.
RGBDComponent	Camera component with configuration for RGB-D input data.

Module Contents

class robokudo.descriptors.camera_configs.components.DepthComponent

Camera component with configuration for depth input data.

topic_depth: str

ROS depth topic name. Must be set by camera configs using this component.

depthOffset: int = 0

Depth offset.

depth_hints: str = 'compressedDepth'

Datatype hint for the depth topic. For example: ‘compressedDepth’ or ‘raw’.

class robokudo.descriptors.camera_configs.components.ColorComponent

Camera component with configuration for color input data.

topic_cam_info: str

ROS camera info topic name. Must be set by camera configs using this component.

topic_color: str

ROS color topic name. Must be set by camera configs using this component.

filterBlurredImages: bool = True

Flag to enable/disable filtering of blurred images

color_hints: str = 'compressed'

Datatype hint for the color topic. For example: ‘compressed’ or ‘raw’.

class robokudo.descriptors.camera_configs.components.TfComponent

Camera component with configuration for transform handling.

tf_from: str

Frame ID of the camera’s optical frame. Must be set by camera configs using this component.

tf_to: str = 'map'

World frame ID to transform the camera’s optical frame to.

lookup_viewpoint: bool = True

Whether to lookup the viewpoint between the camera and the world.

class robokudo.descriptors.camera_configs.components.StableViewpointComponent

Camera component with configuration for viewpoint stabilization.

only_stable_viewpoints: bool = True

Flag to use only stable viewpoints

max_viewpoint_distance: float = 0.01

Maximum allowed distance for viewpoint changes

max_viewpoint_rotation: float = 1.0

Maximum allowed rotation for viewpoint changes

class robokudo.descriptors.camera_configs.components.WorldDescriptorComponent

Camera component with configuration for world descriptors.

world_descriptor: str = 'world_iai_kitchen20'

Filename of the world descriptor configuration

class robokudo.descriptors.camera_configs.components.RGBDComponent

Bases: DepthComponent, ColorComponent

Camera component with configuration for RGB-D input data.

color2depth_ratio: typing_extensions.Tuple[float, float] = (1.0, 1.0)

If the resolution of the depth image differs from the color image, we need to define the factor for (x, y). Example: (0.5,0.5) for a 640x480 depth image compared to a 1280x960 rgb image Otherwise, just put (1,1) here

hi_res_mode: bool = False

Setting this to true will apply some workarounds to match the depth data to RGB on the Kinect.