Detects persons through head and face detection and identifies the detected faces. The results are tracked over time to increase confidence. The system is optimized to be used with a Microsoft Kinect or Asus Xtion Pro Live sensor but could also handle different sensors as long as depth and color images are provided.
face_detector is a package containing algorithms and nodes related to the detection of faces.
See launch/face_dector.wide.launch to launch in continuous detection mode, or launch/face_detector_action.wide.launch to launch as an action.
| classifier_name | A readable string name for the classifier. Will be published with the result. |
| classifier_filename | Full path to the trained haar cascade. Currently useful cascades are haar_frontalface_alt.xml and haar_profileface.xml in opencv2 |
| classifier_reliability | Double 0-1. Some notion of the classifier's reliability for use in a larger system. |
| do_continuous | true = Run continuously. false = Wait for action call |
| do_publish_faces_of_unknown_size | true = 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_display | none = Don't display anything. local = Display in an OpenCV window. |
| face_size_min_m | Double. The minimum width of a face, in meters. Defaults to 0.1m. |
| face_size_max_m | Double. The maximum width of a face, in meters. Defaults to 0.5m. |
| max_face_z_m | Double. 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_m | Double. Only used for tracking. The maximum distance between two face detections before they are considered different faces. Defaults to 1.0m. |