I have a task, I need to make a device that would detect objects in real time (neural network + usb camera) and when an object appears in front of the camera, video recording would turn on, but the work of the neural network would not stop. I tried creating parallel streams 1 stream - neural network, 2 - video recording, but when video recording is turned on, an error occurs that this video stream from this camera is already busy. Are there options to implement this idea?
You can try DeepStream SDK. It is based on gstreamer and you can run pipeline like:
Adding a ghost pad after splitting a pipeline using Tee? - #11 by DaneLLL
And can use tee plugin to simultaneously encode the frames and do object detection.
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.