andrefdre / Dora_the_mug_finder_SAVI

Dora The Mug Finder: Detection and classification of items placed on top of a table using point cloud processing and neural networks.
GNU General Public License v3.0
5 stars 1 forks source link

Create a recording and playing back Kinect data #30

Closed TatianaResend closed 1 year ago

TatianaResend commented 1 year ago

To be able to work at home without the kinect it is necessary to record a bag file.

Calibrated your Kinect (intrinsics and extrinsics):

Recording and playing back Kinect data (http://wiki.ros.org/openni_launch/Tutorials/BagRecordingPlayback):

TatianaResend commented 1 year ago

Start the device:

roslaunch openni_launch openni.launch depth_registration:=true

Record the bag:

rosbag record camera/depth_registered/image_raw camera/depth_registered/camera_info camera/rgb/image_raw camera/rgb/camera_info --duration=20s -O kinect
rosbag info kinect.bag 

Start camera:

roscore
rosparam set /use_sim_time false
roslaunch openni_launch openni.launch load_driver:=false

Open rviz and create a PointCloud2 display of /camera/depth_registered/points, with fixed frame /camera_link.

rosrun rviz rviz  
rostopic hz /camera/depth_registered/points

Playing back data:

rosbag play -l --clock <file_name.bag>
TatianaResend commented 1 year ago

Link to the created bag file: https://uapt33090-my.sharepoint.com/:f:/g/personal/tatianaresende_ua_pt/EhQNmVKEdUdMilgyC_Fa2G0BeYKnZ2Q6EII4-JZgYatFdQ?e=FcT8X1

TatianaResend commented 1 year ago

Link to the bag files "kinect_complete.bag", "kinect_complete_no_chairs", "kinect_table" and the parameters of camera's intrinsic: https://uapt33090-my.sharepoint.com/:f:/g/personal/tatianaresende_ua_pt/EhQNmVKEdUdMilgyC_Fa2G0BeYKnZ2Q6EII4-JZgYatFdQ?e=FcT8X1