Im using TrafficamNet and vehicleTypeNet and it’s work great. I have an rtsp stream with a resolution of 1280x720.
The traffiCamNet model has an input of 960x544
My question is, when cropped images are cascaded over to the vehicleTypeNet classifier, are they being cropped from the 1280x720 rtsp input stream that is set in the stream mux to 1280x720, or is it cropped from the resolution of the trafficamNet model at 960x544?
I would like to understand this so I know what min and max width and height to set for classification of objects.
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
The SGIE will crop the objects images directly from the original video(1280x720) according to the bboxes generated by the PGIE. The gst-nvinfer is open source, it is a “in-place” transform plugin. GstBaseTransform (gstreamer.freedesktop.org)
Please make sure you are familiar with GStreamer before you start with DeepStream.