forked from seanshi007/kinect_based_arm_tracking
-
Notifications
You must be signed in to change notification settings - Fork 0
vkguda355/kinect_based_arm_tracking
Folders and files
Name | Name | Last commit message | Last commit date | |
---|---|---|---|---|
Repository files navigation
Demo Description This is a demo that makes baxter to mimic the arm motion of a human.It use a package, openni_tracker, to get the data of a human body, which is in fact some tf frames related to the camera.Then, it translates these frames to arm angles, which are used to control the baxter. What we need baxter robot(real or in gazebo) microsoft kinect How to use this demo We had some problems in working with the openni in the first place, which showed some errors when we tried to use openni to connect the kinect.Then, following the solution of BastiAB in this website: ros-drivers/openni_tracker#9 (comment) We make it work Here are the steps we take to run the demo: 1.follow the instructions of BastiAB in ros-drivers/openni_tracker#9 (comment) 2.connect the kinect to your PC 3.open a terminal, run: roslaunch openni_launch openni.launch this command will use openni to connect to the kinect then, open a terminal, run: rosrun openni_tracker openni_tracker this command will run the tracker code and thus generate skeletons of human bodies. 4.open a terminal, run: rosrun rviz rviz when rviz is launched, choose "openni_depth_frame" in Fixed Frame, add TF to display. You should see a least one skeleton of a human body now. If nothing shows, make sure some one stands in front of the kinect. 5.choose the index of the body you want to track in rviz. open our puber code, change the value of "user_index" for example: we want to tract the person with 'head_1', then we change the value of "user_index" to be '1' in tf_listen_v8_puber.py now, open a terminal, run: rosrun [package_name] tf_listen_v8_puber this command will make our puber run. the puber is used to receive and smooth the data of the tracked body, then publish them. 6.open a terminal, run: rosrun [package_name] tf_listen_v8_controller this command will make our controller run. it subscribes to the topics of smoothed data published by our puber, and controls the baxter accordingly. 7.now move your arms in front of the kinect, and the baxter will mimic you! Demo Video Link https://www.youtube.com/watch?v=U5lJFFy4i3A Feel free to ask or comment!(this may be a bad "read me", grammatically)
About
Great Kinect arm tracking project with baxter
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published
Languages
- Python 67.4%
- CMake 32.6%