Could not determine type of stream - Deepstream

I’m trying to run an Deepstream App to get YOLO bbox feedback and use that information on a costumized message. I have tested the system using this command
gst-launch-1.0 udpsrc port=5300 ! application/x-rtp, media=video, clock-rate=90000, encoding-name=H264, payload=96 ! rtph264depay ! h264parse ! queue ! h264parse ! video/x-h264 ! avdec_h264 ! nvvideoconvert ! m.sink_0 nvstreammux name=m gpu-id=0 live-source=1 batch-size=1 batched-push-timeout=40000 width=800 height=600 enable-padding=1 nvbuf-memory-type=0 num-surfaces-per-frame=1 ! nvinfer config-file-path=$(pwd)/config_infer_primary_yoloBEV.txt ! nvvidconv ! nvegltransform ! nveglglessink
producing this Pipline GST (35.9 KB).

Using the app, I’m getting the error Could not determine type of stream, respective Pipeline DS (44.9 KB) and deepstream_config.txt (1.4 KB).

What am I missing on configuration file? Or do you suggest other approach?

• Jetson TX2 NX
• DeepStream 6.0.0
• JetPack 4.6
• TensorRT 8.0.1-1+cuda10.2

from the error, it is related to the source, for simplicity, you might use file source instead, for example, gst-launch-1.0 filesrc location=/opt/nvidia/deepstream/deepstream/samples/streams//sample_720p.mp4 ! qtdemux ! h264parse ! nvv4l2decoder ! nvstream …

I have no problem running the gst-launch command that I posted. I’m trying to run the deepstream-app sample, that’s where I have the issue.

uri=udp://@127.0.0.1:5300 is not correct because deepstream-app dose not know other rtp information, deepstream-app can support RTSP source.

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.