Help me, please..

[somyul23]

I want to know how this Deep Learning Streamer Pipeline Server can get sources to detect object from usb-camera-streaming in edgeX RTSP(Real Time Streaming Protocol) server.
Someone said me that RTSP server put the streaming video on the screen and then this microservice cognize the video, check the position and detect objects in the video. If that opinion is right, I think this microservice covers the video like filter. Is that right?
My English is not perfect, so I don't know if the meaning was conveyed well. Answers I'll wait.

[brmarkus]

You might want to have a look into the section "https://github.com/dlstreamer/pipeline-server#rtsp" and try to get it running, to understand how it works.

[nnshah1]

In addition to what @brmarkus mentions in terms of using RTSP as output from the Pipline Server, if you want to consume RTSP streams by the Pipeline Server (i.e. do object detection on an incoming rtsp streams). See examples here:

https://github.com/dlstreamer/pipeline-server/blob/main/docs/customizing_pipeline_requests.md#customizing-pipeline-requests