1、

  1.1 opencv static linux

       mkdir build & cd build

 

  cmake .. -LH  这句话用来查看编译选项  如果不知道编译啥  可以用这个查看一下~

cmake
-D CMAKE_INSTALL_PREFIX=/work/lib/opencv/ubuntu14/2.4.13
-D BUILD_SHARED_LIBS=OFF
-D WITH_CUDA=OFF
-D WITH_OPENCL=OFF
-D BUILD_PERF_TESTS=OFF
-D BUILD_TESTS=OFF
-D BUILD_opencv_world=ON
-D WITH_FFMPEG:BOOL=OFF
-D BUILD_opencv_videoio=OFF
-D BUILD_JPEG=ON
-D BUILD_PNG=ON
..

 

如果遇到ipp 下载出错被Qiang, 则需要手动下载然后配置, 下载地址在cmake 的log 中可以找到

配置位置: opencv/3rdparty/ippicv文件夹下的 ippicv.cmake中,第47行

修改为    "file:///home/ubuntu/Downloads/"

 

 

 注: 如果要加入contrib , 则需要加入如下选项:

-D OPENCV_EXTRA_MODULES_PATH="../../contrib/modules"   即

 

如果要加入libpng libjpeg

-D BUILD_JPEG=ON 
-D BUILD_PNG=ON 

 

 

 

 

 

若需要编译dnn  则需要cmake版本3.5.1 以上, 这里是cmake 3.6.0 下载链接

https://download.csdn.net/download/u011258240/11122952


可能出现的错误:
  1.In-source builds are not allowed : 删掉CMakeCache.txt 然后重新编译
  2. ipp 下载失败 打开buildCMakeDownloadLog.txt 就可以看到下载链接了 下载下来即可
  1.2. opencv + win10
  取消勾选 JAVA python cuda test , 添加 install 安装目录 , configure , generate

安装笔记, caffe 、 opencv等

 

 3.4的依赖

Libs: -L${exec_prefix}/lib/x86_64-linux-gnu

-lopencv_dnn -lopencv_ml -lopencv_objdetect -lopencv_shape -lopencv_stitching -lopencv_superres -lopencv_videostab -lopencv_calib3d -lopencv_features2d -lopencv_highgui -lopencv_videoio -lopencv_imgcodecs -lopencv_video -lopencv_photo -lopencv_imgproc -lopencv_flann -lopencv_core

-L${exec_prefix}/share/OpenCV/3rdparty/lib/x86_64-linux-gnu

-littnotify -llibprotobuf -lzlib -llibjpeg -llibwebp -llibpng -llibtiff -llibjasper -lIlmImf -lippiw -lippicv -ldl -lm -lpthread -lrt -lz



2. caffe

编译机器

1. 安装依赖

sudo apt-get install libprotobuf-dev libleveldb-dev libsnappy-dev libopencv-dev libboost-all-dev libhdf5-serial-dev libgflags-dev libgoogle-glog-dev liblmdb-dev protobuf-compiler  libatlas-base-dev cmake  python-pip

2. 安装  numpy 

pip install numpy

 

3 安装caffe动态库

  下载caffe源码

    wget  https://github.com/BVLC/caffe/archive/1.0.tar.gz

  cd caffe-1.0.0

  cp Makefile.config.example Makefile.config

 

  更改Makefile.config  

第八行   CPU_ONLY := 1  打开,表示只使用CPU,如果你用GPU  这一步就不用做了。

修改 cmake/Dependencies.cmake  

最顶部加入: 

set(CMAKE_PREFIX_PATH ${CMAKE_PREFIX_PATH} "/work/lib/opencv/3.3")
find_package(OpenCV 3.3.0 REQUIRED)

 

mkdir build

cd build

cmake  ../   -DBUILD_SHARED_LIBS=1 -DCMAKE_INSTALL_PREFIX=install

 

  ####  cmake .   -DBUILD_SHARED_LIBS=1 -DCMAKE_INSTALL_PREFIX=install

 

【可选项】修改caffe 源码,使其不要在控制台打印一大堆东西, 将src/caffe/common.cpp  GlobalInit 源码替换为如下

安装笔记, caffe 、 opencv等
void GlobalInit(int* pargc, char*** pargv) {
{
  // Google flags.
  ::gflags::ParseCommandLineFlags(pargc, pargv, true);
  ::google::InitGoogleLogging(*(pargv)[0]);
  google::SetLogDestination(google::WARNING,"");
}
安装笔记, caffe 、 opencv等

 

  make

  make install

 

cd MTCNN_Caffe

cmake .

make

 

运行机器:

apt-get install  libgoogle-glog-dev  libopencv-dev

 

3. CUDA

 3.1cuda sdk

   https://developer.nvidia.com/cuda-toolkit-archive

    1. `sudo dpkg -i cuda-repo-ubuntu1404-10-0-local-10.0.130-410.48_1.0-1_amd64.deb`
    2. `sudo apt-key add /var/cuda-repo-<version>/7fa2af80.pub`
    3. `sudo apt-get update`
    4. `sudo apt-get install cuda`

 其中 Windows版本 VS需要配置 : C:Program FilesNVIDIA GPU Computing ToolkitCUDAv10.1extrasvisual_studio_integrationMSBuildExtensions  复制到 

        $(你的路径)MSBuildMicrosoftVCv160BuildCustomizations

      

3.2  驱动下载

https://www.nvidia.cn/Download/Find.aspx?lang=cn

nvidia-smi #查看当前驱动
ubuntu-drivers devices #查看系统建议安装的驱动
apt-get isntall nvidia-430 #安装驱动

 

可能出现的安装错误: ERROR: Installation has failed. Please see the file ‘/var/log/nvidia-installer.log’ for details. You may find suggestions on fixing installation problems in t

解决方法: https://forums.developer.nvidia.com/t/installing-driver-fails-for-tesla-v100/83983

 

3.3cuDNN  依赖安装

https://developer.nvidia.com/rdp/cudnn-archive

安装cudnn 的时候下载 cudnn library , 不要下载runtime lib  和 dev lib  

 

cuda 和cudnn 的删除   https://blog.csdn.net/wanzhen4330/article/details/81704474

 

3.4  不同版本cuda切换

 ln -sf cuda-8.0/ cuda

 

3.5  不同cudnn版本切换  :    

ln -sf /usr/lib/x86_64-linux-gnu/libcudnn.so.7 /etc/alternatives/libcudnn_so

ln -sf /usr/include/x86_64-linux-gnu/cudnn_v7.h /etc/alternatives/libcudnn

 

3.6   nsight  下载地址

https://developer.nvidia.com/gameworksdownload#?tx=$gameworks,developer_tools

 

 

4. ubuntu14安装tensorflow

#####cd tensorflow-1.13.1

pip install tensorflow

pip install tensorflow-gpu

 

 

 

5 RetinaNet 环境搭建

git clone https://github.com/fizyr/keras-retinanet.git

cd keras-retinanet

pip install numpy --user

pip install . --user

pip install --upgrade Pillow

 

训练: 

####keras_retinanet/bin/train.py pascal /root/darknet/VOCdevkit/VOC2007/

retinanet-train pascal /path/to/VOCdevkit/VOC2007

 

可能出现的错误:

1. Getting error: unknown file type '.pyx' when installing from source #13

解决办法:

sudo apt-get remove python-setuptools
wget https://bootstrap.pypa.io/get-pip.py
sudo -H pip install -U pip setuptools

2. ImportError: libcublas.so.10.0: cannot open shared object file: No such file or directory

解决办法: 安装CUDA 10.0

 

 

6. 搭建Tensorflow1.13   C++ 开发环境

 

  6.1 linux

  https://blog.csdn.net/gubenpeiyuan/article/details/80855644

版本需要按照以下来搭配 , 不然会出问题

https://www.tensorflow.org/install/source

安装笔记, caffe 、 opencv等

192.168.1.208   tensorflow = 1.13.1

 

下载Tensorflow1.3.0  下载 bazel-0.4.5  cudnn6

tar -zxvf v1.3.0.tar.gz

./bazel-0.4.5-installer-linux-x86_64.sh

cd tensorflow*

bazel build --config=opt --config=cuda //tensorflow:libtensorflow_cc.so

 

可能出现的错误和解决方案:

bug: 'protobuf.bzl': no such package ... ... 

fix:      sed -i '@https://github.com/google/protobuf/archive/0b059a3d8a8f8aa40dde7bea55edca4ec5dfea66.tar.gz@d' tensorflow/workspace.bzl

see it : https://github.com/tensorflow/tensorflow/issues/12979

 

 

使用Docker 编译  (refer url: https://www.tensorflow.org/install/source?hl=zh-cn)

Docker 是为 TensorFlow 构建 GPU 支持的最简单方法,因为主机只需安装 NVIDIA® 驱动程序,而不必安装 NVIDIA® CUDA® 工具包。如需设置 nvidia-docker,请参阅 GPU 支持指南和 TensorFlow Docker 指南(仅限 Linux)。

以下示例会下载 TensorFlow :devel-gpu-py3 映像并使用 nvidia-docker 运行支持 GPU 的容器。此开发映像已配置为构建支持 GPU 的 Python 3 pip 软件包:

docker pull tensorflow/tensorflow:devel-gpu-py3
nvidia-docker run -it  -w /tensorflow -v $PWD:/mnt -e HOST_PERMS="$(id -u):$(id -g)" 
    tensorflow/tensorflow:devel-gpu-py3 bash
git pull  # within the container, download the latest source code

注意: 如果docker 不小于19.03 需要使用  ...   docker run --gpus all -it 命令

然后,在该容器的虚拟环境中,构建支持 GPU 的 TensorFlow 软件包:

./configure  # answer prompts or use defaults

#####bazel build --config=opt --config=cuda //tensorflow/tools/pip_package:build_pip_package

  bazel build --config=opt --config=cuda //tensorflow:libtensorflow_cc.so

./bazel-bin/tensorflow/tools/pip_package/build_pip_package /mnt  # create package

chown $HOST_PERMS /mnt/tensorflow-version-tags.whl

注意:1.如果提示bazel版本不匹配,去git下载一个编译好的版本

      2. configure 的编译选项的含义参考 https://blog.csdn.net/yhily2008/article/details/79967118

 

在该容器中安装和验证软件包并检查是否有 GPU:

pip uninstall tensorflow  # remove current version

pip install /mnt/tensorflow-version-tags.whl
cd /tmp  # don't import from source directory
python -c "import tensorflow as tf; print(tf.contrib.eager.num_gpus())"

 

 

 

  6.2 Windows  安装:   (放弃了)

    1) download swig exe  :   http://www.swig.org/

 

  编译成静态库:

  http://www.luohanjie.com/2019-07-17/build-tensorflow-c-static-libraries.html

 

 

 7. bazel   编译 transform_graph    (下载最新版的bazel  和  tensorflow)

  bazel build tensorflow/tools/graph_transforms:transform_graph

 

 

 8  交叉编译opencv

https://docs.opencv.org/2.4/doc/tutorials/introduction/crosscompilation/arm_crosscompile_with_cmake.html

http://bbs.ebaina.com/forum.php?mod=viewthread&tid=38496&highlight=opencv

vi  cmake/OpenCVCompilerOptions.cmake

最末尾添加:

if(ENABLE_NEON)
    add_extra_compiler_option("-mcpu=cortex-a7 -mfpu=neon")
endif()

 

vi CMakeList.txt

将NEON那行改为:

OCV_OPTION(ENABLE_NEON   "Enable NEON instructions"           ON )

 

 

 

 

cmake
-D CMAKE_INSTALL_PREFIX=/work/lib/opencv/arm-hisi/2.4.13
-D BUILD_SHARED_LIBS=ON
-D WITH_CUDA=OFF
-D WITH_OPENCL=OFF
-D BUILD_PERF_TESTS=OFF
-D BUILD_TESTS=OFF
-D BUILD_opencv_world=OFF
-D WITH_FFMPEG:BOOL=OFF
-D BUILD_opencv_videoio=OFF
-D BUILD_JPEG=ON
-D BUILD_PNG=ON
-DCMAKE_TOOLCHAIN_FILE=../platforms/linux/arm-gnueabi.toolchain.cmake
-DCMAKE_CXX_COMPILER=/opt/hisi-linux/x86-arm/arm-hisiv400-linux/bin/arm-hisiv400-linux-gnueabi-g++
-DCMAKE_C_COMPILER=/opt/hisi-linux/x86-arm/arm-hisiv400-linux/bin/arm-hisiv400-linux-gnueabi-gcc
-DCMAKE_AR=/opt/hisi-linux/x86-arm/arm-hisiv400-linux/bin/arm-hisiv400-linux-gnueabi-ar
-DCMAKE_LINKER=/opt/hisi-linux/x86-arm/arm-hisiv400-linux/bin/arm-hisiv400-linux-gnueabi-ld
-DCMAKE_NM=/opt/hisi-linux/x86-arm/arm-hisiv400-linux/bin/arm-hisiv400-linux-gnueabi-nm
-DCMAKE_OBJCOPY=/opt/hisi-linux/x86-arm/arm-hisiv400-linux/bin/arm-hisiv400-linux-gnueabi-objcopy
-DCMAKE_OBJDUMP=/opt/hisi-linux/x86-arm/arm-hisiv400-linux/bin/arm-hisiv400-linux-gnueabi-objdump
-DCMAKE_STRIP=/opt/hisi-linux/x86-arm/arm-hisiv400-linux/bin/arm-hisiv400-linux-gnueabi-strip
.. 

 

 

 

9  openFrameWorks 

https://openframeworks.cc/download/older/

安装0.90

 

10   VS2012   下载地址

http://download.microsoft.com/download/B/0/F/B0F589ED-F1B7-478C-849A-02C8395D0995/VS2012_ULT_chs.iso

产品密钥

  • YKCW6-BPFPF-BT8C9-7DCTH-QXGWC

  • RBCXF-CVBGR-382MK-DFHJ4-C69G8

  • YQ7PR-QTHDM-HCBCV-9GKGG-TB2TM

  • YKCW6-BPFPF-BT8C9-7DCTH-QXGWC

11   glfw-3  安装

isntall  libglfw3

https://launchpad.net/ubuntu/xenial/amd64/libglfw3/3.1.2-3

install libglfw3-dev

https://launchpad.net/ubuntu/yakkety/amd64/libglfw3-dev/3.1.2-3

 

12  VS 旧版本   https://visualstudio.microsoft.com/zh-hans/vs/older-downloads/ 

 

13 docker 安装  pytorch

  https://medium.com/@zaher88abd/pytorch-with-docker-b791edd67850

 

14 docker 查看镜像版本

  去网站 https://hub.docker.com/ 搜索镜像,  点开星星最多的镜像, 在tags 里面搜索符合的镜像  , 然后pull

  docker pull tensorflow/tensorflow:1.15.2-gpu

 

 

 

15  安装Nvidia-docker 

  添加源

curl -s -L https://nvidia.github.io/nvidia-docker/gpgkey | 
  sudo apt-key add -
distribution=$(. /etc/os-release;echo $ID$VERSION_ID)
curl -s -L https://nvidia.github.io/nvidia-docker/$distribution/nvidia-docker.list | 
  sudo tee /etc/apt/sources.list.d/nvidia-docker.list
sudo apt-get update

  安装

sudo apt install nvidia-container-toolkit
systemctl restart docker

 

 

16 intel ipp 下载地址

https://dynamicinstaller.intel.com/system-studio/

https://dynamicinstaller.intel.com/download

 

17 webpp

https://storage.googleapis.com/downloads.webmproject.org/releases/webp/index.html

 

18  libtorch 安装

https://pytorch.org/

 

19 cudnn 安装

https://developer.nvidia.com/rdp/cudnn-download