When I run yolov5 to detect video using the deepstream app, how can I save the target type, confidence, coordinates, and number of frames in each frame of the image to a file？
You can add postprocess by yourself, get the vaules and write in a file.Please refer to the link below:
https://docs.nvidia.com/metropolis/deepstream/dev-guide/text/DS_postprocessing_plugin.html
Can you please go into more detail? I’m new to it, sorry
1.You should learn the basic knowledge of Gstreamer
2.We suggest you learn some structure of the deepstream to know how to get the values you wanted.
https://docs.nvidia.com/metropolis/deepstream/dev-guide/text/DS_plugin_metadata.html
3.You can refer our demo code,such as:
/opt/nvidia/deepstream/deepstream/sources/apps/sample_apps/deepstream-test1/deepstream_test1_app.c
You can learn how to get the values of frame_number, num_rects, vehicle_count, person_count, etc…
4. When you get the values you wanted, you can just write it in a file with any programing language.
Thank you very much for your help
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.