How to use Nvidia DeepStream with Jetson Nano | step by step tutorial
Вставка
- Опубліковано 23 тра 2024
- DeepStream is developed by NVIDIA.
DeepStream is a SDK that includes libraries, APIs, and pre-trained models for building and deploying AI-powered applications.
Email id: aarohisingla1987@gmail.com
Commands to execute :
nvcc --version
This command will display the CUDA toolkit version installed on your system.
dpkg -l | grep TensorRT
This command will display the installed version of TensorRT on your system.
Run the following commands to uninstall DeepStream and its dependencies:
sudo apt-get remove -purge deepstream*
sudo apt-get autoremove
rm -rf ~/.config/deepstream
Installing DeepStream:
step 1:
Add NVIDIA's DeepStream repository:
sudo add-apt-repository universe
sudo add-apt-repository multiverse
sudo add-apt-repository ppa:jonathonf/ffmpeg-4
sudo apt-get update
step 2:
Install Dependencies: Before installing DeepStream, make sure you have all the necessary dependencies. Run the following commands to install them: below mentioned command should install all the necessary dependencies for DeepStream
sudo apt-get install -y \
libssl1.0.0 \
libssl-dev \
libcrypto++6 \
libcrypto++-dev \
libcurl4 \
libopencv-dev \
libcudnn8 \
libnvinfer8 \
libnvinfer-dev \
libnvinfer-plugin8 \
libnvparsers-dev \
libnvparsers8 \
python3 \
python3-pip \
python3-dev \
python3-libnvinfer-dev \
python3-libnvinfer \
python3-numpy \
python3-wheel \
python3-setuptools \
libgstrtspserver-1.0-0 \
Step 3:
Install DeepStream SDK: Now, you can install the DeepStream SDK. Run the following commands:
sudo apt-get install deepstream-6.0
step 4:
Verification: Once installed, you can verify that DeepStream is correctly installed by running:
deepstream-app --version
This should display the version of DeepStream installed on your system.
Next:
cd /opt/nvidia/deepstream/deepstream-6.0/sources/apps/sample_apps
cd deepstream-test1
sudo GST_DEBUG=3 ./deepstream-test1-app /opt/nvidia/deepstream/deepstream-6.0/samples/streams/sample_720p.h264
#nvidia #computervision #objectdetection #jetsonnano - Наука та технологія
Obrigado por postar esse pequeno e extremamente útil tutorial. Era exatamente o que eu precisava ❤
Glad it helped 😊
Mam please continue this one.thank you very much
Sure 👍
Wow! 👏🏻
Thanks!
when i run the command "sudo apt-get install deepstream-6.0"
I get this:
E: Couldn't find any package by regex 'deepstream-6.0'
I'm on a jetson orin nano, i have all dependencies installed too.
Nice tutorial!! but it only worked after following your tutorial on OpenCV with CUDA 😊. The samples applications were not pre-compiled in my case. Thank you
You have to compile the applications when you use it for first time. Sorry, if I missed saying this in my tutorial.
hello i need a question, i'm making project yolov5 on jetson nano and csi camera with realtime detection, and if i tried running config txt for just camera csi it's running, and if i'm added config txt for deepstream like a model engine and will be error, and i'm trying just config txt camera again and will be error. and it have to reboot jetson and tried again running just config txt camera and its running.
can you help me? :')
I need help guy im trying to run a model that will detect a ghost that appear on screen at 45-60fps from roboflow. Which device should buy?
Pls make a video on how to build custom application/projects using DeepStream SDK.
Noted!
It will great and helpful mam if you can tell how to install and activate TPU in Nvdia laptops / one with GPU capability so that we can use that for Jupiter notebook
Noted!
Hey Aarohi... great work... I have been able to implement Deepstream on Jetson Nano thanks to your tutorial... Any idea how to run the Deepstream python apps on the nano? I am not getting any support for that anywhere. Thanks in advance.
Make sure you have Python and the necessary dependencies installed on your Jetson Nano.
Clone the DeepStream Python Apps Repository. Install Required Packages. Then Run DeepStream Python Applications.
@1:14 - Bionic Beaver - didn't know 18.04 was called that. lol
When do we get our Nvidia text-to-3d.
We know they're holding out on us.
The work I could be doing.
The entertainment I could be providing.
Haha, yeah, the codenames are fun. As for Nvidia's text-to-3D, hopefully soon! Can't wait to see what amazing things we can create with it.