Open mohanraja97 opened 6 years ago
Looks like there is no topic /image_raw published. Do you started a camera node? See example here:
$ roscore &
# Launch your preferred usb camera driver
$ rosrun uvc_camera uvc_camera_node &
$ rosrun find_object_2d find_object_2d image:=image_raw
I have searched a lot of packages to publish kinect 360 data, but i couldn't find. Can you suggest any available package that publishes kinect 360 data to the find_object_2d GUI. If possible send me the repository link.
On Wed, 12 Sep 2018 at 00:03, matlabbe notifications@github.com wrote:
Looks like there is no topic /image_raw published. Do you started a camera node? See example here https://github.com/introlab/find-object#run:
$ roscore &
Launch your preferred usb camera driver
$ rosrun uvc_camera uvc_camera_node & $ rosrun find_object_2d find_object_2d image:=image_raw
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/introlab/find-object/issues/67#issuecomment-420375198, or mute the thread https://github.com/notifications/unsubscribe-auth/ApLPNe0cVmK9aK8iUcN3MS6pNXNoYzyjks5uaAHjgaJpZM4Wjr2v .
for my kinect 360 i have been using freenect-glview as my driver, Suggest a package to publish my kinect data
For kinect 360, I usually use freenect_launch:
$ sudo apt-get install ros-kinetic-freenect-launch
$ roslaunch freenect_launch freenect.launch depth_registration:=true
$ roslaunch freenect_launch freenect.launch depth_registration:=true Running the above lanuch node, I am getting an error. The error is no matching device found, you can see this error in the above image's last line.
I do have an other doubt that, do I have to run my driver freenect-glview while running my launch file $ roslaunch freenect_launch freenect.launch depth_registration:=true
If I am running my driver and running my $ roslaunch freenect_launch freenect.launch depth_registration:=true in an other node Iam getting the below error
If you can see the depth and RGB stream with freenect-glview, close it then retry the launch file. This could be a USB problem, is it plugged alone on the USB hub? Try other USB.
What I am getting during my freenect-glview is the below two imgaes The above is the rgb image and the original image
In the above image I am getting only the raw acceleration and the mks acceleration, What does that mean?
How to check for the depth of the image using freenect-glview? With reference to your above comment, I tried using different USB port, But i am the getting the warning as
process[camera_base_link3-24]: started with pid [2458] [ INFO] [1537065507.342738095]: Initializing nodelet with 4 worker threads. [ INFO] [1537065512.354459643]: Number devices connected: 1 [ INFO] [1537065512.354558607]: 1. device on bus 000:00 is a Xbox NUI Camera (2ae) from Microsoft (45e) with serial id 'A00362D00362137A' [ INFO] [1537065512.355636703]: Searching for device with index = 1 [ INFO] [1537065512.738167996]: Starting a 3s RGB and Depth stream flush.
[ WARN] [1537065513.551776475]: Could not find any compatible depth output mode for 1. Falling back to default depth output mode 1. [ INFO] [1537065513.568438067]: rgb_frame_id = 'camera_rgb_optical_frame' [ INFO] [1537065513.568543551]: depth_frame_id = 'camera_depth_optical_frame' [ WARN] [1537065513.588162512]: Camera calibration file /home/mohan/.ros/camera_info/rgb_A00362D00362137A.yaml not found. [ WARN] [1537065513.588392456]: Using default parameters for RGB camera calibration. [ WARN] [1537065513.588480882]: Camera calibration file /home/mohan/.ros/camera_info/depth_A00362D00362137A.yaml not found. [ WARN] [1537065513.588555087]: Using default parameters for IR camera calibration. [camera/camera_nodelet_manager-1] process has died [pid 2435, exit code -11, cmd /opt/ros/kinetic/lib/nodelet/nodelet manager name:=camera_nodelet_manager log:=/home/mohan/.ros/log/7994126a-b959-11e8-bf19-b8763f2d02a3/camera-camera_nodelet_manager-1.log]. log file: /home/mohan/.ros/log/7994126a-b959-11e8-bf19-b8763f2d02a3/camera-camera_nodelet_manager-1*.log ^C[camera_base_link3-24] killing on exit
I think there is no problem with the USB there is some file missing in the repository, see the below image it says process[camera_base_link3-24]: started with pid [11303] [ INFO] [1537068148.195660962]: Searching for device with index = 1 [ INFO] [1537068148.415497446]: Starting a 3s RGB and Depth stream flush.
[ WARN] [1537068149.060112863]: Could not find any compatible depth output mode for 1. Falling back to default depth output mode 1. [ INFO] [1537068150.869693338]: rgb_frame_id = 'camera_rgb_optical_frame' [ INFO] [1537068150.869790812]: depth_frame_id = 'camera_depth_optical_frame' [ WARN] [1537068150.890726583]: Camera calibration file /home/mohan/.ros/camera_info/rgb_A00362D00362137A.yaml not found. [ WARN] [1537068150.890837258]: Using default parameters for RGB camera calibration. [ WARN] [1537068150.890919785]: Camera calibration file /home/mohan/.ros/camera_info/depth_A00362D00362137A.yaml not found. [ WARN] [1537068150.891014682]: Using default parameters for IR camera calibration.
Is depth_image_proc package installed? The freenect-glview seems to show depth image on left. Beware that kinect cannot see depth below ~40cm, it is why the fingers are black (no depth).
You closed the issue, did you make it working?
No I haven't closed my issues blog, that was done by mistake How to change it's state from closed to open?
And I haven't installed depth_image_package. Is it mandatory to install that package? (If yes what does it do)
Can you explain me how these things works and the list of packages that must be installed in order to find the position of the object using Kinect 360 sensor by using ROS Kinetic version?
On Sun, Sep 16, 2018, 10:18 PM matlabbe notifications@github.com wrote:
Is depth_image_proc package installed? The freenect-glview seems to show depth image on left. Beware that kinect cannot see depth below ~40cm, it is why the fingers are black (no depth).
You closed the issue, did you make it working?
— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub https://github.com/introlab/find-object/issues/67#issuecomment-421794461, or mute the thread https://github.com/notifications/unsubscribe-auth/ApLPNdfssxa0r42FrARwXD8VTci7OyVTks5uboDDgaJpZM4Wjr2v .
Normally installing ros-kinetic-freenect-launch
should also install ros-kinetic-depth-image-proc
package. To debug image topics on ROS, use rqt_image_view
and browse published images to see if they are correctly published.
You can also give a try with openni_launch
(ros-kinetic-openni-launch) like in the official example:
http://wiki.ros.org/find_object_2d#A3D_position_of_the_objects
To launch find_object_2d
with only the RGB image after starting a kinect using openni_launch
or freenect_launch
driver:
$ rosrun find_object_2d find_object_2d image:=camera/rgb/image_rect_color
my next step to detect object using find_object_2d is to?