View my account

Intel Real Sense Depth Camera D435i Camera Capabilities

Comments

8 comments

  • Official comment
    Jesus Garcia

    Hello Wanning,

     

    You can use cfg.enable_stream using the index of the stream like this:

     

    cfg.enable_stream(RS2_STREAM_INFRARED, 0);
    cfg.enable_stream(RS2_STREAM_INFRARED, 1);

    Regards,
    Jesus
    Comment actions Permalink
  • Wanninng

    I'm unable to stream the infrared cameras at the moment using my python code, would you know how I would go about that? I'm currently only able to stream the RGB camera and not the infrared. 

    0
    Comment actions Permalink
  • Wanninng

    I decided to use the Intel Sense Viewer to stream the infrared cameras but I want to capture the raw images of the stream. Are you able to provide me a configuration file that would allow me to capture raw images at any given time?

    0
    Comment actions Permalink
  • MartyG

    To get both the left and right IR streams, you should ensure that (1) you are using USB 3 mode and not USB2, and (2) your camera has 'Advanced Mode' enabled.  You can enable Advanced Mode in the RealSense Viewer software.

    If you decide to have another try at doing it in Python, I hope that the script below will be of use to you.

    https://github.com/IntelRealSense/librealsense/issues/3294#issuecomment-466042259

    0
    Comment actions Permalink
  • Wanninng

    Thanks Marty, I currently am able to stream both cameras and when I save the snapshot, I get the files as a PNGs, is there a configuration that would allow me to get .RAW files instead or is running a script the only way to obtain those files?

    0
    Comment actions Permalink
  • MartyG

    When you save a PNG snapshot using the RealSense Viewer, it usually saves a .raw file and a .csv file in the same save location that the PNG is placed in.  The .raw file has an icon that looks like a mountain.

     

    0
    Comment actions Permalink
  • Wanninng

    Didn't know that, thank you very much!

    Lets say I'm running a test where it takes couple hundred photos every minute or so, is there a command line that you can provide that would allow me to change the file names every time it hits snapshot like 1,2,3... etc without someone manually having to type each file name change so it wouldn't over write the previous snapshot inside the SDK software?

    0
    Comment actions Permalink
  • MartyG

    I don't know of a pre-made function like that for saving photos, but ROS supports the specifying of a maximum time length or file size for the recording of a 'bag' file.  When the time limit is reached, ROS closes the recording and automatically starts a new bag file.  Once you have a bunch of bags, you could use a bag conversion tool to convert them into other formats.  The SDK has a 'Convert' tool for doing this.

    https://github.com/IntelRealSense/librealsense/tree/master/tools/convert

    If you do not want to have to compile the tool from source, you can find a pre-built executable under the name rs-convert in the SDK folder RealSense SDK 2.0 > tools

    Details of the ROS file-splitting commands can be found at the link below:

    https://github.com/IntelRealSense/librealsense/issues/2424#issuecomment-423284173

    0
    Comment actions Permalink

Please sign in to leave a comment.