C# wrapper - sample code for object detection in RGB image and pixel position mapping back to point cloud vertex.

[AOndr1]

I'd like to get the color frame from the D415 camera in a form of some basic datatype to e.g. detect circles using c# and then map those circle positions to 3D (to a point cloud vertex).

Could you please point me in the right direction with some sample code?

I'm missing documentation for the c# wrapper.

Thank you.

[MartyG]

The biggest problem is probably the lack of RealSense-compatible object detection sample programs for C#. There are ones for C++ and OpenCV or ROS, and ones for Python, but I have not come across one for C#.. Of the mainstream languages supported by the RealSense SDK (C++, C#, Python), C# tends to have less example programs than the others, unfortunately.

If you can get the 3D depth coordinates of a detected object (RealSense object detection sample programs do offer depth readings of detected objects) then it may be possible to convert them into a 3D point cloud. Again, this would rely on being able to get C# code for such a task.

If you absolutely need to use C#, it would be best to get advice from the RealSense developers at the RealSense GitHub forum by visiting the link below and clicking the 'New Issue' button to post your question there. Good luck!

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

[AOndr1]

Hello MartyG, thank you for the reply - I realized that I didn't make myself quite clear, sorry. What I'm actually interested in is the realsense specific part (and C#) - code/functions that would allow me to get the RGB image (in the form of some basic data-type) and a way to link image coordinates to a pixel in the point cloud.

But nevertheless, thank you for the github tip!