As you know the trained images for the pre-trained TLT models are used normal image. and It’s best to use normal images in the inference phase. but I used 180 degree camera stream for processing.
Q1- I want to know is it possible to use 360d gstreamer plugin for 180 degree image cameras?
Q2- Is it necessary to dwarp 180 degree images for processing?
Q1: If you are talking about nvdwarper plugin Gst-nvdewarper — DeepStream DeepStream Version: 5.0 documentation, please check the configuration of the dewarper whether it can meet your 180 degree camera. Gst-nvdewarper — DeepStream DeepStream Version: 5.0 documentation