How to obtain bounding box coordinated of objects detected in DashCamNet?

Please provide complete information as applicable to your setup.
Jetson Nano
JetPack v4.4
TensorRT v7.1.3
I have downloaded and run dashcamnet on the sample videos provided as well as on my own videos. I would like to get the bounding box coordinates of the objects detected by dashcamnet. Any idea on how to go about this problem?

PS: I am a beginner, please elaborate the answer a little as it might go over my head. sorry for the inconvenience. Thanks in advance.

You can refer to deepstream-app, which provides config property you can set. we have one property to dump primary gie output, including bbox coordinates.
see this, function gie_primary_processing_done_buf_prob in deepstream_app.c


Pathname of an existing directory where the application stores primary detector output in a modified KITTI metadata format

1 Like