1 Reply Latest reply on Jun 5, 2018 5:31 AM by MartyG

    Access D435/R200 Videostream

    NikoBele

      Hello,

       

      i am using C# and the Realsense D435.

      What i am trying to accomplish is to access the D435 Videostream and use the Data to create my own videostream.

      I dont want to grab a frame with var frames = pipeline.WaitForFrames(); but access the raw data and feed my custom videostream with it.

      The idea behind this is to have an application, with an interface to various cameras, for example the Realsense D435, R200, a Kinect or a webcam. All of them using one generic videostream.

      Has someone an idea how to do that, or can someone forward me to a tutorial, or some kind of information?

        • 1. Re: Access D435/R200 Videostream
          MartyG

          It depends on whether you want the stream to be interactive or not.   If you want to just toggle between live videos of what the cameras are seeing, you could use the free video software XSplit Broadcaster.  If each camera was running its own stream-generating application in separate windows, then you could assign each of those windows to a separate XSplit panel and swap between them like toggling the views on a TV sports broadcast.

           

          XSplit can also send the video view you are generating to a browser-based video service such as Twitch to display it live there.  I use Twitch as an example, though I got banned from it once for streaming a video about the virtual world Second Life that was unrelated to videogames, so you should check the terms and conditions of each video service. 

           

          https://www.xsplit.com/#broadcaster

           

          If you want the data to be interactive, that is more difficult because of all the different types of cameras that you want to use.  I would be inclined to use the OpenCV software and its OpenNI compatibility to create a generic stream for your camera types.  D415 / D435 cameras do not have OpenNI support yet though, but Intel apparently have it on their to-do list to look at implementing it.

           

          https://docs.opencv.org/3.2.0/d7/d6f/tutorial_kinect_openni.html