This package uses the data generated by ground_hog, visual_odometry und upper_body_detector to track people. It is also possible to omit the ground_hog detections.
Parameters:
load_params_from_file
default = true:false
tries to read parameters from datacentre,true
reads parameters from YAML file specified byconfig_file
config_file
default = $(find mdl_people_tracker)/config/mdl_people_tracker.yaml: The config file containing all the essential parameters. Only used ifload_params_from_file == true
.machine
default = localhost: Determines on which machine this node should run.user
default = "": The user used for the ssh connection if machine is not localhost.queue_size
default = 20: The synchronisation queue sizecamera_namespace
default = /head_xtion: The camera namespace.rgb_image
default = /rgb/image_rect_color:camera_namespace
+rgb_image
= rgb image topiccamera_info_rgb
default = /rgb/camera_info:camera_namespace
+camera_info_rgb
= rgb camera info topicground_plane
default = /ground_plane: The ground plane published by the upper_body_detectorground_hog
default = /groundHOG/detections: The ground_hog detectionsupper_body_detections
default = /upper_body_detector/detections: The upper_body_detector_detectionsvisual_odometry
default = /visual_odometry/motion_matrix: The visual_odometry datapedestrain_array
default = /mdl_people_tracker/people_array: The resulting tracking data.- `people_markers" default="/mdl_people_tracker/marker_array_: A visualisation array for rviz
- `people_poses" default = /mdl_people_tracker/pose_array_: A PoseArray of the detected people
rosrun:
rosrun mdl_people_tracker mdl_people_tracker [_parameter_name:=value]
roslaunch:
roslaunch mdl_people_tracker mdl_people_tracker_with_HOG.launch [parameter_name:=value]
To run the tracker without the groundHOG feture extraction running use:
rosrun mdl_people_tracker mdl_people_tracker _ground_hog:=""
or
roslaunch mdl_people_tracker mdl_people_tracker.launch [parameter_name:=value]