Module: RGBDSensorToReactive


Module info:

Module name Short description Publishes Subscribes
RGBDSensorToReactive This module implements the adquisition of depth and infrared images from an RGB-D camera (PrimeSense/Asus/Kinect) IRD_WARNING , KINECT1 SHUTDOWN

MOOS commands accepted by the module:

Parameters accepted in the MOOS mission file:
Parameter Description
Count_threshold [if detect_close_obstacles_with_ir=true] Number of pixels with IR value > IR_threshold to send the Reactive Module a warning (something may be close)
detect_close_obstacles_with_ir If enable, points with IR value > IR_threshold will increase the counter of high_ir_px (see Count_threshold param)
discard_high_ir_points If enable, all points with IR value > IR_threshold will be discarded in the depth measurements (to avoid spurious due to light sources)
Downsample Use it to consider coarser images (computationally lighter). It must be integer
floor_limit Points detected with height < floor_limit will be considered as floor and discarded as obstacles
IR_threshold Infrared threshold to consider a point affected by an IR light source
max_depth The max depth (m) the camera is able to detect
min_depth The minimum depth (m) the camera is able to detect
New_sensor Indicates the type of RGBD camera: TRUE=PrimeSense or Asus, FALSE=Kinect
Opengl_scene Whether or not to display an opengl scene with the detected obstacles (used for debug)
pose_pitch The pitch angle (degrees) of the camera on the robot (-80,80) (something logical...)
pose_roll The roll angle (degrees) of the camera on the robot (-90,90)
pose_x The X position (m) of the camera on the robot
pose_y The Y position (m) of the camera on the robot
pose_yaw The yaw angle (degrees) of the camera on the robot (-180,180)
pose_z The Z position (m) of the camera on the robot
Res_height Camera resolution(px) - Shared for depth and infrared images (240 or 480). Only used if New_sensor == true
Res_width Camera resolution(px) - Shared for depth and infrared images (320 or 640). Only used if New_sensor == true

Detailed description:
This module implements the adquisition of depth and infrared images from an RGB-D camera (PrimeSense/Asus/Kinect)and processes these data to send them to the 3D reactive module. The module sorts the points in height bands (according to the height sectionsused to model the robot shape) and finds the most restrictive set of obstacles for each height band. This module should always belaunched together with the "NavigatorReactivePTG3D" because some of its configuration parameters need to be read.

Module graph:
RGBDSensorToReactive

TODO list:
(None)


Changes log:
(None)




Page generated by Mooxygen 1.1.0 at Thu Jan 22 11:30:21 2015
Valid HTML 4.01 Transitional