Jetson Nano

• 使用安裝Jetson Nano with Ubuntu 20.04 OS image

https://github.com/Qengineering/Jetson-Nano-Ubuntu-20-image

• 更改預設python

cd /usr/bin

sudo rm python

sudo ln -s python3.8 python

sudo apt update

pip3 install --upgrade pip

• 增加SWAP

#建立檔案

sudo fallocate -l 8G /var/swapfile

#設定權限

sudo chmod 600 /var/swapfile

#設定交換空間

sudo mkswap /var/swapfile

#啟動交換空間

sudo swapon /var/swapfile

#寫入每次啟用

sudo bash -c 'echo "/var/swapfile swap swap defaults 0 0" >> /etc/fstab'

#驗證是否啟動

sudo swapon --show

sudo free -h

• 確認CUDA

nvcc -V

• 安裝YoloV8

pip install ultralytics

pip uninstall opencv-python

• 安裝OpenCV

pip install numpy==1.23.1

https://qengineering.eu/install-opencv-on-jetson-nano.html

• 測試opencv

>>> import cv2

>>> cv2.__version__

• 檢查有沒有 support CUDA & GStreamer

>>>print(cv2.getBuildInformation())

• 切換開機模式

# 開啟終端機模式

sudo systemctl set-default multi-user.target

# 開啟桌面模式

sudo systemctl set-default graphical.target

# 查看當前的模式

sudo systemctl get-default

• 調整功耗模式

# 鎖住功率使其不過載

sudo jetson_clocks

# 顯示當前模式

sudo nvpmodel -q

# 預設為高效能模式(10W)，此功率需用 DC 5V 4A 供電，不然會突然關機

sudo nvpmodel -m0

# 若使用 Micro-USB 供電需切換到 5W 模式

sudo nvpmodel -m1

• 設定風扇

sudo touch /etc/rc.local

sudo chmod u+x /etc/rc.local

vi /etc/rc.local

寫入rc.local

sleep 10

sudo /usr/bin/jetson_clocks

sudo sh -c 'echo 255 > /sys/devices/pwm-fan/target_pwm'

• 安裝輸入法

$ sudo apt install ibus-pinyin ibus-chewing

#介面新增中文

點選「Setting -> Language Support」-> 「install / remove language」，勾選「Chinese(traditional)」，安裝中文介面，安裝好後「中文(台灣)」拖曳到最上方，選擇「Apply System-Wide」。

#設定輸入法

執行「ibus-setup」，設定輸入法的切換方式，與輸入法選chinese增加chewing(新酷音)

#vi  /etc/network/interfaces

auto wlan0

iface wlan0 inet static

address 192.168.3.241

netmask 255.255.255.0

gateway 192.168.3.254

wpa-ssid [wifi-ssid]

wpa-psk [wifipasswd]

安裝tightvncserver跟xrdp套件，重啟Jetson Nano

sudo apt update

sudo apt install tightvncserver xrdp

sudo reboot

安裝xubuntu-desktop

sudo apt install xubuntu-desktop

echo xfce4-session >~/.xsession

重啟xrdp服務

sudo service xrdp restart

參考資料：

https://blog.cavedu.com/2019/12/19/jetson-nano-remote-desktop-windows-mac-osx/

在/etc/fastab加入下方一行：

//[NAS IP]/[資料夾] /home/[掛載路徑] cifs  username=[帳號],password=[密碼],uid=[uid],gid=[gid],sec=ntlm,user_xattr 0 0

執行sudo mount -a重新執行fstab

• https://github.com/Archiconda/build-tools/releases

• bash Archiconda3-0.2.3-Linux-aarch64.sh

•  echo ". /home/jetbot/archiconda3/etc/profile.d/conda.sh" >> ~/.bashrc

Python套件的AppStore： https://pypi.org/

PS.安裝matplotlib之前，先移除內建的python3-matplotlib：sudo apt remove python3-matplotlib

重要套件：pip3 install -U numpy==1.19.4 protobuf==3.3.0 cython pillow  scipy

                   matplotlib==3.3.2 

機器學習：pip3 install scikit-learn cvzone

網路爬蟲：pip3 install -U requests beautifulsoup4 selenium pandas

若要在虛擬環境下使用上述套件，請在activate檔案中加入：

系統安裝的套件：export PYTHONPATH=/usr/lib/python3.6/dist-packages:$PYTHONPATH

使用者安裝的套件：export PYTHONPATH=/home/jetbot/.local/lib/python3.6/site-packages:$PYTHONPATH

如果用pip安裝opencv-python，無法使用CUDA

• 移除原有的opencv

sudo apt purge *libopencv*

• 檢查是否有移除

pkg-config opencv --modversion

• 顯示CUDA位置

sudo sh -c "echo '/usr/local/cuda/lib64' >> /etc/ld.so.conf.d/nvidia-tegra.conf"

sudo ldconfig

• 安裝第三方套件

sudo apt install build-essential cmake git unzip pkg-config

sudo apt install libjpeg-dev libpng-dev libtiff-dev

sudo apt install libavcodec-dev libavformat-dev libswscale-dev

sudo apt install libgtk2.0-dev libcanberra-gtk* libgtkglextmm-x11-1.2-dev

sudo apt install python3-dev python3-numpy python3-pip

sudo apt install libxvidcore-dev libx264-dev libgtk-3-dev

sudo apt install libtbb2 libtbb-dev libdc1394-22-dev

sudo apt install gstreamer1.0-tools libv4l-dev v4l-utils

sudo apt install libgstreamer1.0-dev libgstreamer-plugins-base1.0-dev

sudo apt install libavresample-dev libvorbis-dev libxine2-dev

sudo apt install libfaac-dev libmp3lame-dev libtheora-dev

sudo apt install libopencore-amrnb-dev libopencore-amrwb-dev

sudo apt install libopenblas-dev libatlas-base-dev libblas-dev

sudo apt install liblapack-dev libeigen3-dev gfortran

sudo apt install libhdf5-dev protobuf-compiler v4l-utils 

sudo apt install libprotobuf-dev libgoogle-glog-dev libgflags-dev

sudo apt install tesseract-ocr libtesseract-dev libleptonica-dev

• 下載並解壓OpenCV

cd ~

wget -O opencv.zip https://github.com/opencv/opencv/archive/4.7.0.zip

wget -O opencv_contrib.zip https://github.com/opencv/opencv_contrib/archive/4.7.0.zip

unzip opencv.zip

unzip opencv_contrib.zip

mv opencv-4.7.0 opencv

mv opencv_contrib-4.7.0 opencv_contrib

rm opencv.zip

rm opencv_contrib.zip

• 建立make

cd ~/opencv

mkdir build

cd build

cmake -D CMAKE_BUILD_TYPE=RELEASE \

-D CMAKE_INSTALL_PREFIX=/usr \

-D OPENCV_EXTRA_MODULES_PATH=~/opencv_contrib/modules \

-D EIGEN_INCLUDE_PATH=/usr/include/eigen3 \

-D WITH_OPENCL=OFF \

-D WITH_CUDA=ON \

-D CUDA_ARCH_BIN=5.3 \

-D CUDA_ARCH_PTX="" \

-D WITH_CUDNN=ON \

-D WITH_CUBLAS=ON \

-D ENABLE_FAST_MATH=ON \

-D CUDA_FAST_MATH=ON \

-D OPENCV_DNN_CUDA=ON \

-D ENABLE_NEON=ON \

-D WITH_QT=OFF \

-D WITH_OPENMP=ON \

-D WITH_OPENGL=ON \

-D BUILD_TIFF=ON \

-D WITH_FFMPEG=ON \

-D WITH_GSTREAMER=ON \

-D BUILD_TESTS=OFF \

-D WITH_EIGEN=ON \

-D WITH_V4L=ON \

-D WITH_LIBV4L=ON \

-D OPENCV_ENABLE_NONFREE=ON \

-D INSTALL_C_EXAMPLES=OFF \

-D INSTALL_PYTHON_EXAMPLES=OFF \

-D PYTHON_EXECUTABLE=/usr/bin/python3.6 \

-D PYTHON3_PACKAGES_PATH=/usr/lib/python3/dist-packages \

-D OPENCV_GENERATE_PKGCONFIG=ON \

-D BUILD_EXAMPLES=OFF ..

• 開始編譯

make -j4

• 安裝

sudo rm -rf /usr/include/opencv4/opencv2

sudo make install

sudo ldconfig

• 測試opencv

>>> import cv2

>>> cv2.__version__

• 檢查有沒有 support CUDA & GStreamer

>>>print(cv2.getBuildInformation())

• 改回dphys-swapfile

• 刪除安裝檔

sudo rm -rf ~/opencv

sudo rm -rf ~/opencv_contrib

• link到虛擬環境

cd ~/[venv_name]/lib/python3.6/site-packages/

ln -s /usr/lib/python3.6/dist-packages/cv2/python-3.6/cv2.cpython-36m-aarch64-linux-gnu.so cv2.so

• https://github.com/Qengineering/Install-OpenCV-Jetson-Nano

更新相依套件

sudo apt-get update

sudo apt-get install git cmake libpython3-dev python3-numpy

下載：

git clone https://github.com/dusty-nv/jetson-inference

cd jetson-inference

git submodule update --init

編譯：

mkdir build

cd build

cmake ..

make

sudo make install

sudo ldconfig

下載模型檔：

cd jetson-inference/tools

./download-models.sh

官方網站：

https://github.com/dusty-nv/jetson-inference/

#安裝套件

$sudo apt install v4l-utils 

#列出攝影機

$v4l2-ctl --list-devices 

#攝影機支援的格式

$v4l2-ctl --list-formats-ext 

• https://makeronsite.com/build_jetbot_env.html

sudo usermod -aG i2c [username]

sudo usermod -aG gpio [username]

sudo chmod 777 /dev/ttyTHS1

• 啟動

cd jetbot/utils

python3 create_stats_service.py

sudo mv jetbot_stats.service /etc/systemd/system/jetbot_stats.service

sudo systemctl enable jetbot_stats

sudo systemctl start jetbot_stats

• 關閉

sudo systemctl disable jetbot_stats

1. 更新 pip3 並安裝 jupyter lab

pip install --upgrade pip

pip install jupyter jupyterlab

sudo reboot

2. 生成 jupyter lab 配置文件

jupyter lab --generate-config

vi ~/.jupyter/jupyter_lab_config.py

修改

c.ServerApp.allow_origin = '*'

c.ServerApp.ip = '0.0.0.0'

3. 設置 jupyter notebook 密碼

jupyter-lab password

設置完後，密碼會存到

~/.jupyter/jupyter_server_config.json

4. 設定開機自動啟動 jupyter lab，創建 jupyter.service 文件

sudo vi /etc/systemd/system/jupyter.service

加入

[Unit]

Description=Jupyter Notebook

[Service]

Type=simple

User=[username]

ExecStart=/home/[username]/.local/bin/jupyter-lab --port 8888 --no-browser

WorkingDirectory=/home/[username]/

[Install]

WantedBy=default.target

5.運行 jupyter

執行以下命令啟動服務

• 自動啟動

sudo systemctl enable jupyter

sudo systemctl start jupyter

#關閉：sudo systemctl disable jupyter

• 手動啟動

~/.local/bin/jupyter-lab --port 8888 --no-browser

檢查服務是否有運行

sudo systemctl status jupyter

6.加入虛擬環境到jupyter

安裝ipykernel

pip3 insall ipykernel

加入虛擬環境

 sudo python -m ipykernel install --name <venv_name>

git clone https://github.com/NVIDIA-AI-IOT/jetcam

cd jetcam

sudo python3 setup.py install

https://github.com/NVIDIA-AI-IOT/jetcam

sudo visudo

%sudo   ALL=(ALL:ALL) NOPASSWD:ALL

查看SD卡掛載在哪裡

lsblk

備份前先sync

sync

安裝dc3dd

sudo apt install dc3dd

將SD卡寫成img檔

sudo dc3dd if=/dev/mmcblk0 of=[USB掛載路徑]/myimg.img

https://github.com/jkjung-avt/tensorrt_demos

https://github.com/linghu8812/tensorrt_inference/tree/master/project/Yolov4

• NVidia官方Jetson Nano資料

• Nvidia Jetbot資料

• Getting Started With Jetson Nano

• DLI Getting Started with AI on Jetson Nano

• https://github.com/dusty-nv/jetson-inference

• Jetson Community Projects：https://developer.nvidia.com/embedded/community/jetson-projects

• Jetson_Zoo：https://elinux.org/Jetson_Zoo

• Cavedu技術部落格

• 映像檔燒錄工具

• 初學 Jetson Nano 不說 No - CAVEDU 教你一次懂

• CIRCUS Pi Jetson Nano 系列

• NVIDIA Jetson Nano 系列文章