face_detector Documentation


Face detection in images.

face_detector is a package containing algorithms and nodes related to the detection of faces.


See the launch files for examples. launch/face_dector.<camera>.launch launches in continuous detection mode, and launch/face_detector_action.<camera>.launch launches as an action.

classifier_nameA readable string name for the classifier. Will be published with the result.
classifier_filenameFull path to the trained haar cascade. Currently useful cascades are haar_frontalface_alt.xml and haar_profileface.xml in opencv2
classifier_reliabilityDouble 0-1. Some notion of the classifier's reliability for use in a larger system.
do_continuoustrue = Run continuously. false = Wait for action call.
do_publish_faces_of_unknown_sizetrue = If depth info is not available, publish the resulting face with a position of (0,0,0). false = Don't publish faces if stereo information isn't available.
do_displayfalse = Don't display anything. true = Display detections in an OpenCV highgui window.
face_size_min_mDouble. The minimum width of a face, in meters. Defaults to 0.1m.
face_size_max_mDouble. The maximum width of a face, in meters. Defaults to 0.5m.
max_face_z_mDouble. The maximum distance of a face from the camera, in meters. (In the camera frame, depth is along the z-axis.) Defaults to 8.0m.
face_separation_dist_mDouble. Only used for tracking. The maximum distance between two face detections before they are considered different faces. Defaults to 1.0m.
use_rgbdDouble. true = use data from an RGBD camera (like the Kinect), false = use stereo data
approximate_sync.true = use data from unsynchronized image-depth pairs (Kinect), false = use data from synchronized image-depth pairs (stereo)

Code API

Author(s): Caroline Pantofaru
autogenerated on Sun Feb 21 2021 03:56:45