How can I customize the output msg and image/video from gst-nvinfer?

Accelerated Computing Intelligent Video Analytics DeepStream SDK
1

Question 1:

For example, if the Primary GIE has detected a car in the frame, how can I send a message(like bbox etc.) with a encoded jpeg to a server? I guess it can be done by NvDsBatchMeta with nvmsgconv?

Question 2:

How can we save the part of video when a car was detected by Primary GIE, and send to a server with a message attached?

Thanks!!

2

Hi,

1. You can design your own metadata message.
Please check this sample for the information:
/opt/nvidia/deepstream/deepstream-4.0/sources/apps/sample_apps/deepstream-user-metadata-test

2. Are you trying to save the output of Deepstream into a video?
If yes, you can just update the sink component with type=3.
https://docs.nvidia.com/metropolis/deepstream/dev-guide/index.html#page/DeepStream_Development_Guide%2Fdeepstream_app_config.3.2.html%23wwpID0ENHA

Thanks.