Table of Contents
- Robot person-following system with multiple sensors (usable with SOBIT EDU and SOBIT PRO)
$ cd ~/catkin_ws/src/
$ git clone https://github.com/TeamSOBITS/sobit_follower
$ cd sobit_follower
# Install the necessary packages for follow me
$ bash install.sh
# Setup the installed package, then catkin_make
$ cd ~/catkin_ws
$ catkin_make- 2D Point Cloud Leg Detection with DR-SPAAM
- GitHub：Person Detection in 2D Range Data modified to work with Python3
- Used in Multiple Sensor Person Tracking
- For more information
- Kalman filter library with two observables as input
- Can also work with a single observation
- Equation of state is a constant velocity model
- Used in Multiple Sensor Person Tracking
- For more information
- Person tracking using a 2D-LiDAR sensor combined with an RGB-D sensor on a pan-tilt rotation mechanism
- Person tracking using 2D point cloud leg detection with DR-SPAAM and person detection with SSD
- For more information
- Person-following control using a Virtual Spring Model with obstacle avoidance by Dynamic Window Approach
- For more information
- Person-following control using Multiple Sensor Person Tracking and Person Following Control
- Users can activate person-following by launching this package's Launch
- Shell scripts are also available to acquire rosbags for experiments and plot the acquired data
- For more information
- Person-following control by Multiple Sensor Person Tracking and Person Following Control using SOBIT EDU
- path：
sobit_follower/launch/sobit_edu/sobit_edu_follower_me.launch - For more information
$ roslaunch sobit_follower sobit_edu_follower_me.launch rviz:=false rqt_reconfigure:=false use_rotate:=true use_smoother:=true
# Arguments
# rviz : whether to start Rviz (bool)
# rqt_reconfigure : whether to start rqt_reconfigure (bool)
# use_rotate : activate SensorRotator (bool)
# use_smoother : whether to perform velocity smoothing (bool)-
SOBIT_EDU enables the combination of the target identification method person_identification_nodelet with SOBIT_EDU to identify the target of the pursuit. Person-following driving made possible by combining the target identification method
-
path：
sobit_follower/launch/sobit_edu/sobit_edu_follower_me_id.launch
$ roslaunch sobit_follower sobit_edu_follower_me_id.launch rviz:=false rqt_reconfigure:=false use_rotate:=true use_smoother:=true
# Arguments
# rviz : whether to start Rviz (bool)
# rqt_reconfigure : whether to start rqt_reconfigure (bool)
# use_rotate : activate SensorRotator (bool)
# use_smoother : whether to perform velocity smoothing (bool)- Person-following control by Multiple Sensor Person Tracking and Person Following Control using SOBIT PRO
- path：
sobit_follower/launch/sobit_pro/sobit_pro_follower_me.launch - For more information
$ roslaunch sobit_follower sobit_pro_follower_me.launch rviz:=false rqt_reconfigure:=false use_rotate:=true use_smoother:=true
# Arguments
# rviz : whether to start Rviz (bool)
# rqt_reconfigure : whether to start rqt_reconfigure (bool)
# use_rotate : activate SensorRotator (bool)
# use_smoother : whether to perform velocity smoothing (bool)- path：
sobit_follower/launch/sobit_pro/sobit_pro_follower_me_id.launch - For more information
$ roslaunch sobit_follower sobit_edu_follower_me_id.launch rviz:=false rqt_reconfigure:=false use_rotate:=true use_smoother:=true
# Arguments
# rviz : whether to start Rviz (bool)
# rqt_reconfigure : whether to start rqt_reconfigure (bool)
# use_rotate : activate SensorRotator (bool)
# use_smoother : whether to perform velocity smoothing (bool)- ssd_pose_ros.launch.xml
- RGB image-based person detector
- path：
sobit_follower/launch/include/ssd_pose_ros.launch.xml - For more information
- dr_spaam_ros.launch.xml
- 2D LiDAR-based person detector
- path：
sobit_follower/launch/include/dr_spaam_ros.launch.xml - For more information
- person_id.launch.xml
- Target identification method using RGB-D sensors
- path：
sobit_follower/launch/include/person_id.launch.xml - For more information
- sobit_edu_tracker.launch.xml
- 2D-LiDAR sensor with SOBIT EDU combined with RGB-D sensor on pan-tilt rotation mechanism for person tracking
- path：
sobit_follower/launch/include/sobit_edu/sobit_edu_tracker.launch.xml - For more information
- sobit_pro_tracker.launch.xml
- 2D-LiDAR sensor with SOBIT PRO combined with RGB-D sensor on pan-tilt rotation mechanism for person tracking
- path：
sobit_follower/launch/include/sobit_pro/sobit_pro_tracker.launch.xml - For more information
- sobit_edu_person_following_control.launch.xml
- Driving control that incorporates obstacle avoidance using the Dynamic Window Approach into tracking control using a Virtual Spring Model with SOBIT EDU
- path：
sobit_follower/launch/include/sobit_edu/sobit_edu_person_following_control.launch.xml - For more information
- sobit_pro_person_following_control.launch.xml
- Driving control that incorporates obstacle avoidance using the Dynamic Window Approach into tracking control using a Virtual Spring Model with SOBIT PRO
- path：
sobit_follower/launch/include/sobit_pro/sobit_pro_person_following_control.launch.xml - For more information
- tracker_param.yaml
- Parameters for person tracking
- path：
sobit_follower/param/tracker_param.launch.xml - For more information
- ssd_param.yaml
- Parameters for RGB image-based person detector
- path：
sobit_follower/param/ssd_param.launch.xml - For more information
- dr_spaam_param.yaml
- Parameters for 2D LiDAR-based person detector
- path：
sobit_follower/param/dr_spaam_param.launch.xml - For more information
- sensor_rotator_param.yaml
- Parameters for pan-tilt rotation control of RGB-D sensor
- path：
sobit_follower/param/sensor_rotator_param.launch.xml
- following_control_param.yaml
- Parameters for driving control
- path：
sobit_follower/param/following_control_param.launch.xml - For more information
- velocity_smoother_param.yaml
- Parameters for speed smoothing
- path：
sobit_follower/param/velocity_smoother_param.launch.xml - For more information
- OSS
- Improved documentation
- Unified coding style
See the open issues for a full list of proposed features (and known issues).