Recording videos with voice by using D455

• MartyG

  • November 21, 2021 07:21
  • Edited

  Hi Ahmdqassabashi  If you are aiming to record to non-interactive video, the easiest way to do it may be to set up the three cameras in the RealSense Viewer tool so that depth and RGB streams from all three cameras are displayed in individual panels that are all visible onscreen at the same time.

  Then you could use video capture software such as XSplit Broadcaster to capture the full screen whilst recording a live voiceover from an external microphone at the same time (RealSense 400 Series cameras do not have a built-in mic).  Further information about this can be found at the link below.

  https://github.com/IntelRealSense/librealsense/issues/6841

  The skeleton part of the project could be a problem though if using the above method, because if the cameras are streaming depth and RGB then they would likely be inaccessible to a separate skeleton tracking program such as Nuitrack SDK or Cubemos Skeleton Tracking SDK, preventing skeleton data from being recorded at the same time.

  It is possible that you could write your own multiple camera application that could perform all of these tasks simultaneously and then screen-capture it so that mics could be recorded over the top of it in real-time.  It would be a complex project to create though.

  Also, you would need a computer with a powerful processor such as Intel Core i7 to stream all of this data from one computer with three cameras and record simultaneously

  0

  Comment actions Permalink
• Ahmdqassabashi

  • November 21, 2021 15:15

  Thank you MartyG;

  Actually, I need to save each video  as a separated file (3 cameras record 6 videos (3 RGB videos and 3 depth videos compatible with the RGB videos ) synchronously and the output is 6 videos ), but your way means that I will record the screen which will contain all the videos. 

  please, can you tell me how I can do that?

  Also, for skeleton video, I think that after the record of the RGB videos, is there any programs which can convert RGB videos to Skeleton videos?

  0

  Comment actions Permalink
• MartyX Grover

  • November 21, 2021 16:00
  • Edited

  What you are aiming for is technically possible.  For example, the link below has a Python script for saving depth and color separately as .avi format video files using the OpenCV VideoWriter instruction.

  https://github.com/IntelRealSense/librealsense/issues/2731#issuecomment-529938267

  To save pairs of .avi from the three different cameras though, you may have to define three separate pipelines and assign each camera to one of the pipelines, accessing the streams via the unique serial number of the camera.  A Python example of defining a set of streams for separate pipelines is below.

  https://github.com/IntelRealSense/librealsense/issues/2882

   

  An example of a skeleton tracking application that can use an .avi video file as input and convert it into a skeleton video is OpenPose

  https://github.com/CMU-Perceptual-Computing-Lab/openpose

  OpenPose is a C++ application, though it has a Python API.

  https://github.com/CMU-Perceptual-Computing-Lab/openpose/blob/master/doc/03_python_api.md

  0

  Comment actions Permalink

Please sign in to leave a comment.