Yolo-FastestV2配置环境

配置环境

Debian换源

修改配置文件

sudo nano /etc/apt/sources.list

内容修改为

deb https://mirrors.tuna.tsinghua.edu.cn/debian/ stretch main contrib non-free
deb-src https://mirrors.tuna.tsinghua.edu.cn/debian/ stretch main contrib non-free
deb https://mirrors.tuna.tsinghua.edu.cn/debian/ stretch-updates main contrib non-free
deb-src https://mirrors.tuna.tsinghua.edu.cn/debian/ stretch-updates main contrib non-free
deb https://mirrors.tuna.tsinghua.edu.cn/debian/ stretch-backports main contrib non-free
deb-src https://mirrors.tuna.tsinghua.edu.cn/debian/ stretch-backports main contrib non-free
deb https://mirrors.tuna.tsinghua.edu.cn/debian-security/ stretch/updates main contrib non-free
deb-src https://mirrors.tuna.tsinghua.edu.cn/debian-security/ stretch/updates main contrib non-free

执行更新

sudo apt update && sudo apt upgrade

NCNN

环境配置

sudo apt-get install -y gfortran
sudo apt-get install -y libprotobuf-dev libleveldb-dev libsnappy-dev libopencv-dev libhdf5-serial-dev protobuf-compiler libprotoc-dev
sudo apt-get install --no-install-recommends libboost-all-dev
sudo apt-get install -y libgflags-dev libgoogle-glog-dev liblmdb-dev libatlas-base-dev

获取ncnn

cd ncnn
git clone https://github.com/Tencent/ncnn

修改CMakeList.txt

将

add_subdirectory(examples)
add_subdirectory(benchmark)

前的注释删除

编译安装ncnn

cd build
mkdir build
cmake ../
make -j4
make install 

Opencv安装

安装python & pip

sudo apt install python3 -y
sudo apt install -y python3-pip

配置pip

pip config set global.index-url https://pypi.tuna.tsinghua.edu.cn/simple 

pip3 install --upgrade pip

安装opencv

pip3 install opencv-python opencv-contrib-python numpy matplotlib opencv-python-headless
pip3 install onnxruntime  onnx-simplifier onnx skl2onnx

验证OpenCV安装

# python3
> import cv2
> cv2.__version__

ncnn环境变量

sudo cp ~/ncnn/build/install/include/ncnn/* /usr/include/
sudo mkdir -p /usr/local/lib/ncnn/
sudo cp ncnn/build/src/libncnn.a /usr/local/lib/ncnn/

/etc/profile写入

export PATH=$PATH:$HOME/bin
C_INCLUDE_PATH=/usr/include/:/usr/include/opencv4:$C_INCLUDE_PATH
export C_INCLUDE_PATH
CPLUS_INCLUDE_PATH=$CPLUS_INCLUDE_PATH:/usr/include/:/usr/include/opencv4
export CPLUS_INCLUDE_PATH
LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/include/:/usr/include/opencv4
export LD_LIBRARY_PATH
LIBRARY_PATH=$LIBRARY_PATH:/usr/include/:/usr/include/opencv4
export LIBRARY_PATH

测试ncnn

前提

• 64位树莓派系统
• NCNN框架
• 64位opencv
• 安装codeblocks sudo apt-get install codeblocks

获取yolofast2-yolo for RPi

wget https://github.com/Qengineering/YoloFastestV2-ncnn-Raspberry-Pi-4/archive/refs/heads/main.zip

or

git clone https://github.com/Qengineering/YoloFastestV2-ncnn-Raspberry-Pi-4

解压，使用codeblocks打开YoloFastestV2.cbp，编译运行

安装yolo fastest

环境配置

sudo apt install git
git clone https://github.com/dog-qiuqiu/Yolo-FastestV2.git
#失败时可以改为git clone git://github.com/dog-qiuqiu/Yolo-FastestV2.git

• armv7l

pip install torch torchvision torchsummary -f https://torch.maku.ml/whl/stable.html
pip install -r Yolo-FastestV2\requirements.txt

• others

pip install -r Yolo-FastestV2\requirements.txt

验证安装

cd Yolo-FastestV2
python3 test.py --data data/coco.data --weights modelzoo/coco2017-epoch-0.235624ap-model.pth --img img/000230.jpg