deepstream-test2_Jetson_Nano

Accelerated Computing Intelligent Video Analytics DeepStream SDK
1

Hi everyone, I am starting to read about DeepStream and I have runned the script example for “ deepstream-test 2 ” on Jetson Nano.
I have tryed to use file and rtsp such as input and i have the same error.

ERROR from element file-source: Resource not found.
Error details: gstfilesrc.c(533): gst_file_src_start (): /GstPipeline:dstest2-pipeline/GstFileSrc:file-source:
No such file “file:/opt/nvidia/deepstream/deepstream-5.0/samples/streams/cars.mp4”
Returned, stopping playback
Deleting pipeline

I have all the links for “model-file”, “proto-file”, etc for the “dstest2_pgie_config.txt”, “dstest2_sgie1_config.txt”, etc files according to Primary and Secondaries paths on Deepstream folder for the samples models.

I am interesting to know the performance for the NVTRACK in order to count the objects.

Thanks a lot !!!

3

Can you provide platform and software information in following format:

• Hardware Platform (Jetson / GPU)
• DeepStream Version
• JetPack Version (valid for Jetson only)
• TensorRT Version
• NVIDIA GPU Driver Version (valid for GPU only)

Are you using DeepStream python or DeepStream C/C++?
deepstream-test 2 is open source for both c/C++ and python.
Have you read README file in the folder /opt/nvidia/deepstream/deepstream-5.0/sources/apps/sample_apps/deepstream-test2(for C/C++)?
It can only accept h264 element stream local file as input.

4

Hi @Fiona.Chen

• Hardware Platform (Jetson / GPU) Jeston Nano

• DeepStream Version 5.0

• JetPack Version (valid for Jetson only) 4.4

• TensorRT Version 7.1.3

• NVIDIA GPU Driver Version (valid for GPU only)

—Yes, I have but I do not understand when it say “It can only accept h264 element stream local file as input.” I have tried with “file:…/xxx.mp4” “rtsp://…”

—I prefer to use Python, I have downloaded the repository and I have write the inks for the files, engine, protox, etc on each config file (dstest2_pgie_config.txt, dstest2_sgie1_config.txt, etc

—My intention is try to understand how the NVTRACKER works for count the objects detected on the scene but not for each frame, I am talking about count the objects which are on the scene since it appears until it disappears but only once.

5

Are you running “deepstream_test_2.py”? deepstream_python_apps/apps/deepstream-test2 at master · NVIDIA-AI-IOT/deepstream_python_apps · GitHub

If so, what is your command line?
The README of python deepstream-test2 is here: deepstream_python_apps/README at master · NVIDIA-AI-IOT/deepstream_python_apps · GitHub

You reported the “Resource not found” problem in this topic, do you want to resolve this problem?

If you have special question for “nvdstracker”, you need to create a new topic. https://docs.nvidia.com/metropolis/deepstream/dev-guide/index.html#page/DeepStream%20Plugins%20Development%20Guide/deepstream_plugin_details.html#wwpID0E0540HA

It is better to learn DeepStream before you start to learn DeepStream python.
https://docs.nvidia.com/metropolis/deepstream/dev-guide/index.html#page/DeepStream_Development_Guide/deepstream_quick_start.html#wwpID0E0VF0HA

6

Thanks @Fiona.Chen , I will take in count the advices, to be honest I am confusing because some texts recommend such as first step run the examples for learn and understand how Deepstream works.