Hello. I was given the task of creating a device that accepts 4 streams (mipi-csi2) and displays in various modes on the screen (similar to video surveillance systems). Picture-in-picture, tile, etc. mode The main thing is the minimum delay. I tested jetson nano with imx219 camera got a delay of 100ms. Question: by connecting 4 cameras, how can I group them and display on the screen which tools to use. (Gstreamer, deepstream … ???) and can I get a delay of up to 100ms.
You can reference to the multi-session of the argus_camera.
This sample aggregate 6 to one display.