Questions on 360 Degree Smart Parking Application

Hi team,

I pulled perception deepstream app (deepstream_360d:4.0.1-19.11) and run well. It is a very useful reference.

  1. I am interested in your implementation of those dependent plugins (GST-NVBBOXFILTER, GST-NVAISLE and GST-NVSPOT) but I cannot find those sources codes in the docker. Appreciated if you let me know where the codes are so that I can study on.

  2. For multi-camera tracking module, I am not sure about your settings of those 360d cameras for this demo. In your github, it said “in your garage setting, you have around 150 cameras – each camera covering a few 10s of meters of aisle”. But in your perception docker demo, the provided videos seem were all taken from an indoor parking lot instead of outdoor. May I know the actual settings of your cameras used in the demo of perception app docker? What’s the topology relationship between those cameras used in those videos?

Thanks in advance for your kind help!

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU): dGPU
• DeepStream Version:4.0.2
• JetPack Version (valid for Jetson only)
• TensorRT Version: 6.0.1.5
• NVIDIA GPU Driver Version (valid for GPU only): 435

please note that the multi-camera tracking module works irrespective of indoor or outdoor camera setup. It is primarily driven by correct calibration, how camera / image coordinates are mapped real world coordinates


In the above blog a set of indoor cameras were used