Use gst-nvdsmetamux to fuse 2 obj ids meta from nvpreprocessed src

NVIDIA-AI-IOT / deepstream_parallel_inference_app

A project demonstrating how to use nvmetamux to run multiple models in parallel.

87 stars 21 forks source link

Use gst-nvdsmetamux to fuse 2 obj ids meta from nvpreprocessed src #2

Open OsamaRyees opened 1 year ago

OsamaRyees commented 1 year ago

Hi, I have split a single cam source into 4 different ROIs and feed it to inference layer. This solve issue of detecting small objects but How to use gst-nvdsmetamux to fuse adjacent objects from different ROI into one object to track