Gstreamer vs fisheye and multiple output accessing

I’m have two problem need to be cleared as following:

  • The first, I have an gstreamer pipeline that working properly for reading rtsp: gst-launch-1.0 rtspsrc location=<location> latency=200 ! rtph264depay ! h264parse ! omxh264enc ! nvvidconv ! video/x-raw, format=BGRx ! videoconvert ! xvimagesink, but when I replace location string by a fisheye ip camera, it didn’t work. What did I do wrong?

  • The second, If I want to get 2 output from the same input by applying gstreamer tee element (ie, once is original and the other is resized frame), how can I save and access them in C++/Python code? Any example or document can help me in this task?
    Sorry for my bad English :3

You can use gstreamer via the C-API or C++ bindings or even the Python bindings (probably easier to start).

From NVIDIA side there are alse some C/C++ wrappers to write plugins for GStreamer (easy type image in/out,…). I think mainly via deepstream.

BR, Markus