Handle streams with different resolutions

[Shannon_Govekar]

I have a couple of cameras with resolution of 720 (height) x 1280(width) and a couple of cameras with the opposite solution i.e 1280 (height) x 720 (width). My application currently sets the streamuxer width and height to a fixed value but that can be changed my reading the input stream height and width.
My problem is, how should my tracker config and config_nvdsanalytics file (where I set my region of interest) handle resolution differences in streams?
Also, does the Deepstream analytics toolkit support relative coordinates instead of absolute? In that case a difference in streaming resolutions shouldn’t cause any problems.

• Hardware Platform (Jetson / GPU) GPU
• DeepStream Version 6.0.1
• JetPack Version (valid for Jetson only)
• TensorRT Version
• NVIDIA GPU Driver Version (valid for GPU only) 470.63
• Issue Type( questions, new requirements, bugs)
• How to reproduce the issue ? (This is for bugs. Including which sample app is using, the configuration files content, the command line used and other details for reproducing)
• Requirement details( This is for new requirement. Including the module name-for which plugin or for which sample application, the function description)

[kesong]

You can configure nvdsanalytics based on stream id. Please refer below document.

https://docs.nvidia.com/metropolis/deepstream/dev-guide/text/DS_plugin_gst-nvdsanalytics.html

[Shannon_Govekar]

What about the config-width and config-height for which the analytics rules are configured? Can I configure these values for different streams?

In one config file, I’d like to feed in multiple cameras ROIs with different resolutions

[property]
enable=1
#Width height used for configuration to which below configs are configured
config-width=720
config-height=1280
osd-mode=2
#Set OSD font size that has to be displayed
display-font-size=12

[Shannon_Govekar]

Oh I think I understand this better now, I just want to confirm.

For the nvdsanalytics: the config-width and config-height have to stay constant and I add ROI info for different resolution streams using the config-width and config-height as my constant reference?

For streammux in order to handle different resolutions I need to use the export USE_NEW_NVSTREAMMUX which doesn’t require me to set output muxer width and height? What about the batch timeout? How would that be calculated? based on incoming frequencies?

I see this error when using the new streammux export flag once the program ends or application stops, not hurting atm… just informing.

[ERROR push 317] push failed [-2]
[ERROR push 317] push failed [-2]
[ERROR push 317] push failed [-2]
[ERROR push 317] push failed [-2]

[kesong]

The new streammux can’t support input with different resolutions. Please check heterogeneous batching in below:

https://docs.nvidia.com/metropolis/deepstream/dev-guide/text/DS_plugin_gst-nvstreammux2.html#heterogeneous-batching

[Shannon_Govekar]

So the new streammux and the old streammux don’t support different resolutions? I know that the old streammux supports scaling however I noticed for the old streammux, if my stream input resolutions and output resolution (streammux) are not the same inference is affected.

Also, when should I use the new streammux over the older one?

[kesong]

There is no update from you for a period, assuming this is not an issue anymore. Hence we are closing this topic. If need further support, please open a new one. Thanks

New streammux support different resolution is in our roadmap. Please wait for our future release. Thanks.