periphery's profile - activity

Hi,

I am using RGBDSLAM on Ubuntu 11.10, on ROS Electric, and a Kinect camera. My goal is to basically move around my room with the camera and create a 3-D map of my room. I guess that just means turning up the RGBDSLAM GUI on the terminal, and then hitting space to record data from the video stream generated. I am wondering, after this, I can simply save these data, but do I have to do anything further? Say, for instance, write some code to further process this data? Or that is enough for the sole purpose I am trying to achieve?

Thanks.

@Felix Endres : I couldn't find the use_glwidget parameter in any launchfile. Which launchfile were you talking about specifically? I am in ros/rgbdslam/launch directory.

Just wondering what the GPU requirements for RGBDSLAM are. I am using a pretty mediocre graphics card, and RGBDSLAM seems to very laggy and in general unable to display images properly, so I don't know if it might be an issues with the graphics card.

Hi, I am using RGBDSLAM on a Ubuntu 11.10 machine, with ROS Electric installed. I am using a Kinect camera to get the video feed. When I open the rgbdslam GUI using:

rosrun rgbdslam rgbdslam

I can see the monochrome image and depth image, but on the top window, I am supposed to see the 'map', right? But what I see is basically the background wallpaper (the desktop) right through it. I am confused as to why this happens.

I also ran rviz, and so far it does say that I am subscribed to a pointcloud2. However, even in rviz I do not get to see a 'map' or a video feed, just a black background screen as set by default.

Can anyone help?

Hi, I am using ROS Diamondback on Ubuntu 10.04 running on VMware Player. I have also connected a Kinect camera via USB. I think I have installed all required drivers and build all the necessary files without failures.

However, when I execute this command:

roslaunch rgbdslam kinect+rgbdslam.launch

The RGBDSLAM GUI opens up, but it doesn't display any video. Two windows in the bottom say "Waiting for monochrome image..." and "Waiting for depth image..."

When I press space and stop, I see that 0 bytes of data has been recorded. The same thing happens when I press enter to get a single frame.

I am sort of stuck at a dead end, and really have no clue where things are going wrong. Any help would be appreciated.

Can anyone verify whether rviz absolutely doesn't work on a virtual machine or not? I am using ROS Diamondback and a Kinect camera (whose drivers have been installed properly. Thanks.

EDIT Hi, I am using VMware Player, and my local machine is Windows 7 (64 bit).

I will put some more details about how I am trying to hook up Kinect. I am following instructions from this link: http://www.ros.org/wiki/openni_camera

To be specific, I am having issues with Step 4 in the link above. For example, for this command:

roslaunch openni_camera openni_node.launch

The terminal window notifies that is looking for a device, and when I plug in my Kinect, it detects it. So I can assume that this is working?

I then do: rosrun rviz rviz And the rviz GUI opens up. I set up and PointCloud 2 display and change all the parameters as mentioned, even changing the background to grey, but then the screen just shows grey colour, and no video feed from the Kinect camera. And the status is "Warning". And the .Global Status is "OK".

Sorry for a very long and confusing post. But I am really new to ROS and I have had a lot of trouble trying to configure SLAM in it.

Thanks for reading.