Contributing to Holscan SDK or holohub

Healthcare & Life Sciences Holoscan SDK
1

I have been developing an application using holoscan SDK. I have implemented certain operators.

  1. Operator to read frames from a video directly using ffmpeg.
  2. Preprocessing operator to process image using some inbuild function and also allowing user to write his own preprocessing function in python.

Also created an application to test multiple models at once directly using the above operators with some user defined configurations.

How can I add these functionalities in holoscan sdk. What is the standard procedure to contribute to holoscan sdk repo with new operators and sample application?

2

This is great - here are some details on contributing your work to the broader community:

holohub/CONTRIBUTING.md at main · nvidia-holoscan/holohub · GitHub

3

Hi @aquraini, I read the guidelines. I have written all operators in python for my application. Do I need to write those operators separately in holohub/operators/ or can I use put all those operators directly in my application in holohub/applications/.
My python operators can be imported directly in python application. Do I need to add support for my python operators for cmake to build them?

4

Hi there! It would be great if you could add those operators to holohub/operators/ so they can be re-used by other applications, since putting those operators directly in your application in holohub/applications/ will only let your operators be used in your application. For reference on a Python operator please see holohub/operators/webrtc_client at main · nvidia-holoscan/holohub · GitHub and holohub/operators/webrtc_server at main · nvidia-holoscan/holohub · GitHub, and the applications that make use of them at holohub/applications/webrtc_video_client at main · nvidia-holoscan/holohub · GitHub and holohub/applications/webrtc_video_server at main · nvidia-holoscan/holohub · GitHub.