ubuntu系统人工智能框架安装指南及常见问题解决方案——numpy、sklearn、TensorFlow、opencv、caffe2、caffe及中间的相关插曲

Table of Contents

1 申明

2 相关框架的安装

2.1 准备工作

2.2 numpy

2.3 sklearn

2.4 TensorFlow

2.5 opencv

2.6 caffe2（纯cpu版）

2.7 caffe(重头戏）

2.7.1 现有的相关配置（因机而异）

2.7.2 需要的配置（基于现有配置和其相互关联）

2.7.3 安装NVIDIA GPU驱动

2.7.4 安装CUDA

2.7.5 配置cuDNN

2.7.6 安装并配置Caffe

3 小结 && 常见问题 && 参考网址

3.1 内存不足

3.2 循环登录

3.3 两个变量的替换

3.3.1 CV_LOAD_IMAGE_COLOR

3.3.2 CV_LOAD_IMAGE_GRAYSCALE

3.4 nvcc fatal : Unknown option ‘fPIC’

3.5 gflags安装的动态链接库没有设置成fPIC

3.6 不同gcc版本的下载和切换

3.6.1 安装gcc

3.6.2 切换gcc

3.7 the ISO C++ 2011错误

3.8 CUDA driver version is insufficient for CUDA runtime version

3.9 Linux下系统自带python和Anaconda切换

3.10 安装opencv的借鉴原文

3.11 安装CUDA、cuDNN的借鉴原文

3.12 安装caffe的借鉴原文

3.13 安装NVIDIA的借鉴原文

3.14 python2 手动安装更新pip

3.15 Python- 解决PIP下载安装速度慢

3.16 关于在ubuntu平台下使用apt-get命令下载速度太慢的问题解决

3.17 Downloading ippicv_linux_20151201.tgz...超时

3.18 Downloading vgg_generated_120.i ... 超时

3.19 Cmake编译opencv源码错误

3.20 git clone速度太慢解决方案

3.21 ImportError:No module named skimage.io

3.22 error while loading shared libraries: libcudnn.so.7: cannot open shared object file: No such file or

3.23 import caffe失败 No module named caffe

3.24 caffe2安装教程的借鉴原文

---

 

1 申明

ubuntu16.04（建议装双系统，不建议虚拟机，因虚拟机无法更改主机显卡驱动，影响caffe的安装），其上已自带python2.7和python3.5

下面安装中的pip默认是对python2.7的安装，若欲安装在python3.5上，将pip换成pip3即可

注释：opencv和caffe无论python2还是python3都可以import。

所需文件绝大部分均可百度云上获取。

百度云地址：

链接：https://pan.baidu.com/s/1MjiFqAgeNNIyCAg5DqDUqA 
提取码：w816 

2 相关框架的安装

2.1 准备工作

注释：linux安装多用sudo，可以少很多麻烦事……

sudo apt-get update
sudo pip install --upgrade pip

pip升级了，但是新的pip版本在install其他包的时候，会有

from pip import main ImportError: cannot import name 'main'

一类的错误，解决方案是:

pip文件在目录下，cd进去

sudo gedit /usr/bin/pip

进行以下修改——

把下面这行

from pip import main

改成

from pip._internal import main

再装个git，方便下东西

sudo apt-get install git

再装一个CMake，后面用得到

sudo apt install cmake

 

2.2 numpy

提示：在安装大型AI框架的时候最好先安装numpy，因为不少框架都依赖它。

一行代码安装：

sudo pip install numpy

2.3 sklearn

又是一行代码安装

sudo pip install sklearn

为了更快乐的生活（画图），可以顺便把matplotlib也装上的

sudo pip install matplotlib

或者

sudo apt-get install python-matplotlib

2.4 TensorFlow

还是一行代码的事

sudo pip install tensorflow

GPU好的小伙伴可以安装其gpu版本

sudo pip install tensorflow-gpu

2.5 opencv

剩下这几个框架是重头戏了，开始详细讲解安装过程。

1 简便安装

pip install opencv-python
pip install opencv-contrib-python

2 源码安装

先构建工具

sudo apt-get install build-essential

然后是必须包

sudo apt-get install cmake git libgtk2.0-dev pkg-config libavcodec-dev libavformat-dev libswscale-dev

下面这些包是可选的（韩信点兵，多多益善）

sudo apt-get install python-dev python-numpy libtbb2 libtbb-dev libjpeg-dev libpng-dev libtiff-dev 

然后下载opencv和opencv的contrib库（保存在~目录）（opencv2）

git clone https://github.com/Itseez/opencv.git
git clone https://github.com/Itseez/opencv_contrib.git

或者（推荐opencv3）

wget https://github.com/opencv/opencv_contrib/archive/3.2.0.zip
wget https://github.com/opencv/opencv/archive/3.2.0.zip

接着进入~/opencv目录

cd opencv

创建release文件夹并进入其

mkdir release && cd release

然后使用cmake，根据我们的配置对opencv进行编译，如下

cmake -D CMAKE_BUILD_TYPE=RELEASE -D CMAKE_INSTALL_PREFIX=/usr/local -D CUDA_GENERATION=Kepler -D OPENCV_EXTRA_MODULES_PATH=~/opencv_contrib/modules ~/opencv/

可以看到，结尾的~/opencv_contrib/modules和~/opencv/应该是git下载下来的opencv_contrib和opencv的对应位置，如果不同应对应修改。特别的，其中的“-D CUDA_GENERATION=Kepler”可以解决编译 opencv 的错误：Unsupported gpu architecture 'compute_20'

这需要等待一小会儿……

然后在~/opencv/release目录下

sudo make

等待编译。（不要小看这个make，可能需要一个多小时都不一定，这取决于您的计算机的处理能力）

等不及的可以使用多线程，再其后加上“-jX”（用X个线程）或“-j”（用所有线程），例如

sudo make -j4

最好根据自己的内存来，不然后报内存溢出错误（欲速则不达）。个人建议，一个线程赋予2g的内存量比较稳妥。我是一根8g的内存条，所以用的4个线程。

 

中间可能会报错

Unable to import the yaml module

解决方案：

sudo pip install pyyaml

再重新make等待至100%，然后安装

sudo make install

最后检查版本（是个X.X.X就说明没问题）

pkg-config --modversion opencv

 

2.6 caffe2（纯cpu版）

先安装依赖项：

$ sudo apt-get update
$ sudo apt-get install -y --no-install-recommends \
    build-essential \
    cmake \
    git \
    libgoogle-glog-dev \
    libgtest-dev \
    libiomp-dev \
    libleveldb-dev \
    liblmdb-dev \
    libopencv-dev \
    libopenmpi-dev \
    libsnappy-dev \
    libprotobuf-dev \
    openmpi-bin \
    openmpi-doc \
    protobuf-compiler \
    python-dev \
    python-pip                          
$ sudo pip install \
    future \
    numpy \
    protobuf
$ sudo apt-get install -y --no-install-recommends libgflags-dev

然后下载pytorch（caffe现已整合在其中）

git clone --recursive https://github.com/pytorch/pytorch.git && cd pytorch

更新一波submodule

git submodule update --init

进入pytorch文件夹

cd pytorch

创建build文件夹并进入

mkdir build && cd build

然后编译安装（又是一个多小时……）

cmake ..

中间可能会报错

Unable to import the `typing` module

解决方案

sudo pip install typing

然后回去重复执行cmake，再等到100%，最后安装

sudo make install

然后验证一波安装

python2 -c 'from caffe2.python import core' 2>/dev/null && echo "Success" || echo "Failure"

结果是Success就对了

python2 -c 'from caffe2.python import workspace; print(workspace.NumCudaDevices())'

结果是0，那是因为这是cpu版本的不是gpu版本的所以能跑就无所谓了。

最后再完善一下环境变量，sudo打开bashrc

sudo gedit ~/.bashrc

任意空行处添加：

export PYTHONPATH=/usr/local:$PYTHONPATH
export PYTHONPATH=~/pytorch/build:$PYTHONPATH
export LD_LIBRARY_PATH=/usr/local/lib:$LD_LIBRARY_PATH

注意其中的PYTHONPATH=~/pytorch/build如果和你的pytorch目录不同的话，要对应修改。

然后使其生效即可

source ~/.bashrc

最后的快乐生活如下

zjhao666@ubuntu:~$ python2
Python 2.7.12 (default, Nov 12 2018, 14:36:49) 
[GCC 5.4.0 20160609] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import numpy
>>> import sklearn
>>> import tensorflow as tf
>>> import cv2
>>> import caffe2
>>> 

到此则证明以上框架已经全部正确安装完毕

2.7 caffe(重头戏）

2.7.1 现有的相关配置（因机而异）

Ubuntu 16.04 x64操作系统， GPU：GM620（集成显卡），MX150（独立显卡）。值得注意的是，这里我选择了集成显卡（应该是错误的），但根据后文来看，CUDA读取的实际是独立显卡，但巧合的是GM620的配置对于MX150同样适用（不影响后文的应用展示）。

之前安装的OpenCV 3.2.0

2.7.2 需要的配置（基于现有配置和其相互关联）

NVIDIA 显卡驱动（需正确适应GPU)

CUDA 9.0（较为稳定的版本）

cnDNN v7.4.2 （符合CUDA 9.0，但不是唯一）

所以，接下来的工作是

1. 安装NVIDIA GPU 驱动
2. 安装CUDA
3. 配置cuDNN
4. 安装并配置Caffe

2.7.3 安装NVIDIA GPU驱动

有个暴力安装版本的，愿意等的也可以，这个成功率极高。

sudo add-apt-repository ppa:graphics-drivers/ppa
sudo apt-get update
sudo apt-get install nvidia-390 #此处数字要对应上面查询到的版本号
sudo apt-get install mesa-common-dev
sudo reboot

或者更暴力的

sudo ubuntu-drivers autoinstall

安装完显卡驱动后，系统需要重启加载驱动，注意如果按照上述流程进行驱动安装的同学，那在重启系统时，会出现一个蓝色背景的界面 perform mok management , 本人在这一步选择了 continue reboot， 导致新安装的 N 卡驱动没有加载，正确的做法如下： 
1. 当进入蓝色背景的界面perform mok management 后，选择 enroll mok , 
2. 进入enroll mok 界面，选择 continue , 
3. 进入enroll the key 界面，选择 yes , 
4. 接下来输入你在安装驱动时输入的密码， 
5. 之后会跳到蓝色背景的界面perform mok management 选择第一个 reboot

这样，重启后N卡驱动就加载了，恭喜你，Ubuntu 18.04LST 安装成功。

不愿意等的，可以走下面。

首先，去https://www.geforce.cn/drivers（推荐，因网站反应速度较快），或NVIDIA官网http://www.nvidia.com/Download/index.aspx?lang=en-us，查看适合自己显卡的驱动并下载。

如下图所示

点击“开始搜索”，结果如下

推荐第一个，因此我们下载“NVIDIA-Linux-x86_64-390.87”，点击下载图标即可进入下载界面，进行下载。

得到的文件为“NVIDIA-Linux-x86_64-390.87.run”。

首先删除掉电脑上可能存在的其他NVIDIA驱动。

进入文本模式，按键：

CTRL+ALT+F1 

接着执行删除操作

sudo apt-get remove nvidia-*
sudo apt-get autoremove

然后通过run文件删除NVIDIA驱动

sudo nvidia-uninstall

接着，重新安装驱动。

进入NVIDIA-Linux-x86_64-390.87.run所在的文件目录，

sudo service lightdm stop
sudo chmod a+x ./NVIDIA-Linux-x86_64-390.87.run
sudo ./NVIDIA-Linux-x86_64-390.87.run -no-x-check -no-nouveau-check -no-opengl-files

参数说明：

 -no-x-check安装驱动时关闭x服务;
 -no-nouveau-check 安装驱动时禁用Nouveau
 -no-opengl-files 安装时只装驱动文件，不安装Opengl

然后重启lightdm

sudo service lightdm restart

这样就实现了NVIDIA GPU驱动的安装，且不会出现循环登录的问题。

重新启动电脑后，登录账号进入推行界面后，输入

sudo nvidia-smi

若如下图所示

则说明NVIDIA已经正常安装。

2.7.4 安装CUDA

1 下载CUDA

登录官网 https://developer.nvidia.com/cuda-90-download-archive （CUDA9.0）下载。特别地，要注意CUDA和NVIDIA显卡驱动的适配性。

现在的情况是：

1. CUDA_8.0支持375.**及以上系列的显卡驱动；
2. CUDA_9.0支持384.**及以上系列的显卡驱动；
3. CUDA_9.1支持389.**及以上系列的显卡驱动。

鉴于我们之前下载了NVIDIA-Linux-x86_64-390.87，点击我们配置的对应图标，所以这里我们选择CUDA9.0（进入的也正是9.0的官网）的对应版本进行下载。

2 运行安装程序

进入cuda_9.0.176_384.81_linux.run所在的目录后（我就在~下），

执行以下命令

sudo chmod 777 cuda_9.0.176_384.81_linux.run
sudo ./cuda_9.0.176_384.81_linux.run

注意：执行后会有一系列提示让你确认，但是注意，有个让你选择是否安装nvidia384驱动时，一定要选择否：Install NVIDIA Accelerated Graphics Driver for Linux-x86_64 384.**? 因为前面我们已经安装了更加新的nvidia384，所以这里不要选择安装。其余的都直接默认或者选择是即可。

3 环境变量设置

打开~/.bashrc文件

sudo gedit ~/.bashrc

将以下内容写入到~/.bashrc尾部：

export PATH=/usr/local/cuda-9.0/bin${PATH:+:${PATH}}
export LD_LIBRARY_PATH=/usr/local/cuda/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}

4 测试CUDA是否安装成功

执行以下命令

cd /usr/local/cuda-9.0/samples/1_Utilities/deviceQuery
sudo make
sudo ./deviceQuery

如果显示一些关于GPU的信息，则说明安装成功。如下图所示：

2.7.5 配置cuDNN

cuDNN是GPU加速计算深层神经网络的库。

1 首先去官网下载cuDNN

cuDNN官网 https://developer.nvidia.com/rdp/cudnn-download，需要注册一个账号才能下载（免费的）。下载cuDNN时也一定要注意与CUDA版本的适配性。

cuDNN在持续更新中，只要对应（我们这里是CUDA9.0）就好了，因此不一定要最新的。我在这里选择的版本是cudnn-9.0-linux-x64-v7.4.2.24.solitairetheme8，即7.4.2版本，不是最新的，但同样可用。

2 进入cuDNN目录后进行解压

sudo tar -xzvf cudnn-9.0-linux-x64-v7.4.2.24.solitairetheme8

得到cuda文件夹

3 进入解压后的文件夹下的include目录

cd cuda/include
sudo cp cudnn.h /usr/local/cuda/include

4 进入lib64目录下，对动态文件进行复制和软链接

 

cd ..
cd lib64
sudo cp lib* /usr/local/cuda/lib64/
cd /usr/local/cuda/lib64/
sudo rm -rf libcudnn.so libcudnn.so.7
sudo ln -s libcudnn.so.7.4.2 libcudnn.so.7
sudo ln -s libcudnn.so.7 libcudnn.so

2.7.6 安装并配置Caffe

1 安装相关依赖

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

2 下载Caffe

git命令行下载

git clone https://github.com/BVLC/caffe.git

或在git网站 https://github.com/BVLC/caffe 上，下载zip后用unzip命令解压。

unzip caffe-master.zip
sudo mv caffe-master caffe
cd caffe

3 修改Makefile.config

因为make指令只能make Makefile.config文件，而Makefile.config.example是caffe给出的makefile例子。因此，首先将Makefile.config.example的内容复制到Makefile.config：

sudo cpMakefile.config.example Makefile.config

打开Makefile.config文件

sudo gedit Makefile.config

并根据个人情况，这里以我为例，作如下修改：

（1）使用了cudnn，则将

#USE_CUDNN:= 1

修改成：

USE_CUDNN:= 1

（2）使用了opencv，且版本是3的，则将

# USE_OPENCV := 0

改成

 USE_OPENCV := 1

，并把

# OPENCV_VERSION:= 3

改成

OPENCV_VERSION:= 3

（3）要使用python来编写layer，则将

# WITH_PYTHON_LAYER:= 1 

改为

WITH_PYTHON_LAYER := 1

（4）重要的一项，将

#Whatever else you find you need goes here. 

下面的

INCLUDE_DIRS:= $(PYTHON_INCLUDE) /usr/local/include
LIBRARY_DIRS:= $(PYTHON_LIB) /usr/local/lib /usr/lib

INCLUDE_DIRS:= $(PYTHON_INCLUDE) /usr/local/include /usr/include/hdf5/serial
LIBRARY_DIRS:= $(PYTHON_LIB) /usr/local/lib /usr/lib /usr/lib/x86_64-linux-gnu /usr/lib/x86_64-linux-gnu/hdf5/serial

这是因为Ubuntu16.04的文件包含位置发生了变化，尤其是需要用到的hdf5的位置，所以需要更改这一路径。

4 修改Makefile文件

打开Makefile文件，做如下修改，将：

NVCCFLAGS +=-ccbin=$(CXX) -Xcompiler -fPIC $(COMMON_FLAGS)

改为

NVCCFLAGS += -D_FORCE_INLINES -ccbin=$(CXX) -Xcompiler -fPIC $(COMMON_FLAGS)

特别提醒， -Xcompiler前有一个空格，不少博客没有该空格会导致后期的错误。

5 编译

make all -j4

同前文一样，一个线程给2g内存比较稳妥。否则中途很可能会报错退出，欲速则不达。

6 测试

sudo make runtest

正确结果如下

到此，caffe框架已被正确安装。

 

3 小结 && 常见问题 && 参考网址

3.1 内存不足

有一个奇怪的bug，当我同时make opencv和caffe两个框架的时候，出现了部分字段如下

c : internal compiler error: 段错误 (program cc1plus) Please submit a full

的bug，查看原因是因为内存不足，所以请分开安装，等待编译和安装的时候可以干其他事情。

3.2 循环登录

安装显卡驱动失败后，导致Ubuntu处于循环登录状态，解决方案可参考：

http://www.cnblogs.com/talugirl/p/5870875.html

或

https://blog.csdn.net/emiedon/article/details/79589526

提示：这两者都失败也不妨多百度或者谷歌，重装系统代价太大了。

3.3 两个变量的替换

来自以下两个编译时的错误

3.3.1 CV_LOAD_IMAGE_COLOR

make error:/caffe2/caffe2/image/image_input_op.h:277:20: error: ‘CV_LOAD_IMAGE_COLOR’ was not declared in this scope #755

将 CV_LOAD_IMAGE_COLOR 替换为 cv::IMREAD_COLOR 或 IMREAD_COLOR

参考网址

https://github.com/facebookarchive/caffe2/issues/755

3.3.2 CV_LOAD_IMAGE_GRAYSCALE

Traceback (most recent call last):
File "/Users/n1/Desktop/FaceDetection/face.py", line 8, in 
 gray = imread(fname, CV_LOAD_IMAGE_GRAYSCALE )
NameError: name 'CV_LOAD_IMAGE_GRAYSCALE' is not defined

将 CV_LOAD_IMAGE_GRAYSCALE 替换为 cv::ImreadModes::IMREAD_GRAYSCALE

参考网址

https://stackoverflow.com/questions/27424285/cv-load-image-grayscale-is-not-definedpy

这两个问题应该是opencv2和3的版本差异导致的。

3.4 nvcc fatal : Unknown option ‘fPIC’

其实之前我提到过，是-Xcompiler之前少了一个空格，即

NVCCFLAGS += -D_FORCE_INLINES -ccbin=$(CXX)-Xcompiler -fPIC $(COMMON_FLAGS)  

正确应为

NVCCFLAGS += -D_FORCE_INLINES -ccbin=$(CXX) -Xcompiler -fPIC $(COMMON_FLAGS)  

参考网址

https://blog.csdn.net/u013524303/article/details/81609643

3.5 gflags安装的动态链接库没有设置成fPIC

编译过程中在make的时候出现下面的错误。

/usr/bin/ld: /usr/local/lib/libgflags.a(gflags.cc.o): relocation R_X86_64_32 against `.rodata.str1.1' can not be used when making a shared object; recompile with -fPIC
/usr/local/lib/libgflags.a: 无法添加符号: 错误的值

参考网址

https://blog.csdn.net/xiaolangwj/article/details/79100667

3.6 不同gcc版本的下载和切换

因为有的软件版本需要高版本或者低版本的gcc或者g++，以支持正常的编译，所以

3.6.1 安装gcc

参考网址

https://blog.csdn.net/weixin_35762621/article/details/80336291

3.6.2 切换gcc

参考网址

https://blog.csdn.net/lyy14011305/article/details/63685779?utm_source=blogxgwz3

3.7 the ISO C++ 2011错误

#error This file requires compiler and library support for the ISO C++ 2011

参考网址

https://blog.csdn.net/jay463261929/article/details/59591104

3.8 CUDA driver version is insufficient for CUDA runtime version

参考网址

https://blog.csdn.net/nijun1992/article/details/80446226

3.9 Linux下系统自带python和Anaconda切换

参考网址

https://blog.csdn.net/zhangxinyu11021130/article/details/64125058

3.10 安装opencv的借鉴原文

参考网址

http://embedonix.com/articles/image-processing/installing-opencv-3-1-0-on-ubuntu/

3.11 安装CUDA、cuDNN的借鉴原文

参考网址

https://blog.csdn.net/u010801439/article/details/80483036

3.12 安装caffe的借鉴原文

参考网址

https://blog.csdn.net/yggaoeecs/article/details/79163789

https://blog.csdn.net/qq_35451572/article/details/79743476

3.13 安装NVIDIA的借鉴原文

参考网址

https://blog.csdn.net/linhai1028/article/details/79445722

3.14 python2 手动安装更新pip

参考网址

https://blog.csdn.net/qiang12qiang12/article/details/76216399

要才能下的资源，开头的百度云网盘里也有的。

3.15 Python- 解决PIP下载安装速度慢

参考网址

https://blog.csdn.net/qq_16481211/article/details/81081996

3.16 关于在ubuntu平台下使用apt-get命令下载速度太慢的问题解决

参考网址

https://www.cnblogs.com/tademeng/p/7451125.html

3.17 Downloading ippicv_linux_20151201.tgz...超时

方法参考网址

https://blog.csdn.net/fjx812/article/details/54946021

ippicv_linux_20151201.tgz免费下载地址

参考网址

https://github.com/opencv/opencv_3rdparty/blob/ippicv/master_20151201/ippicv/ippicv_linux_20151201.tgz

3.18 Downloading vgg_generated_120.i ... 超时

方法

openc3.2进行cmake时需要下载的文件。 下载vgg_generated_120.i放在opencv_contrib/modules/xfeatures2d/cmake/.download/e8d0dcd54d1bcfdc29203d011a797179。

下载地址

见开头的百度云盘（csdn上c币太贵了）。

3.19 Cmake编译opencv源码错误

CMake Error: The following variables are used in this project, but they are set to NOTFOUND.
Please set them or make sure they are set and tested correctly in the CMake files:
CUDA_nppi_LIBRARY (ADVANCED)
linked by target "opencv_cudev" in directory D:/Cproject/opencv/opencv/sources/modules/cudev
linked by target "opencv_cudev" in directory D:/Cproject/opencv/opencv/sources/modules/cudev
linked by target "opencv_test_cudev" in directory D:/Cproject/opencv/opencv/sources/modules/cudev/test
linked by target "opencv_core" in directory D:/Cproject/opencv/opencv/sources/modules/core
linked by target "opencv_core" in directory D:/Cproject/opencv/opencv/sources/modules/core
linked by target "opencv_test_core" in directory D:/Cproject/opencv/opencv/sources/modules/core
linked by target "opencv_perf_core" in directory D:/Cproject/opencv/opencv/sources/modules/core
linked by target "opencv_test_cudaarithm" in directory D:/Cproject/opencv/opencv/sources/modules/cudaarithm
linked by target "opencv_cudaarithm" in directory D:/Cproject/opencv/opencv/sources/modules/cudaarithm
linked by target "opencv_cudaarithm" in directory D:/Cproject/opencv/opencv/sources/modules/cudaarithm
linked by target "opencv_perf_cudaarithm" in directory D:/Cproject/opencv/opencv/sources/modules/cudaarithm
linked by target "opencv_flann" in directory D:/Cproject/opencv/opencv/sources/modules/flann
linked by target "opencv_flann" in directory D:/Cproject/opencv/opencv/sources/modules/flann
linked by target "opencv_test_flann" in directory D:/Cproject/opencv/opencv/sources/modules/flann
linked by target "opencv_imgproc" in directory D:/Cproject/opencv/opencv/sources/modules/imgproc
linked by target "opencv_imgproc" in directory D:/Cproject/opencv/opencv/sources/modules/imgproc
linked by target "opencv_test_imgproc" in directory D:/Cproject/opencv/opencv/sources/modules/imgproc
linked by target "opencv_perf_imgproc" in directory D:/Cproject/opencv/opencv/sources/modules/imgproc
linked by target "opencv_test_ml" in directory D:/Cproject/opencv/opencv/sources/modules/ml
linked by target "opencv_ml" in directory D:/Cproject/opencv/opencv/sources/modules/ml
linked by target "opencv_ml" in directory D:/Cproject/opencv/opencv/sources/modules/ml
linked by target "opencv_test_video" in directory D:/Cproject/opencv/opencv/sources/modules/video
linked by target "opencv_video" in directory D:/Cproject/opencv/opencv/sources/modules/video
linked by target "opencv_video" in directory D:/Cproject/opencv/opencv/sources/modules/video
linked by target "opencv_perf_video" in directory D:/Cproject/opencv/opencv/sources/modules/video
linked by target "opencv_cudabgsegm" in directory D:/Cproject/opencv/opencv/sources/modules/cudabgsegm
linked by target "opencv_cudabgsegm" in directory D:/Cproject/opencv/opencv/sources/modules/cudabgsegm
linked by target "opencv_test_cudabgsegm" in directory D:/Cproject/opencv/opencv/sources/modules/cudabgsegm
linked by target "opencv_perf_cudabgsegm" in directory D:/Cproject/opencv/opencv/sources/modules/cudabgsegm
linked by target "opencv_cudafilters" in directory D:/Cproject/opencv/opencv/sources/modules/cudafilters
linked by target "opencv_cudafilters" in directory D:/Cproject/opencv/opencv/sources/modules/cudafilters
linked by target "opencv_test_cudafilters" in directory D:/Cproject/opencv/opencv/sources/modules/cudafilters
linked by target "opencv_perf_cudafilters" in directory D:/Cproject/opencv/opencv/sources/modules/cudafilters
linked by target "opencv_cudaimgproc" in directory D:/Cproject/opencv/opencv/sources/modules/cudaimgproc
linked by target "opencv_cudaimgproc" in directory D:/Cproject/opencv/opencv/sources/modules/cudaimgproc
linked by target "opencv_test_cudaimgproc" in directory D:/Cproject/opencv/opencv/sources/modules/cudaimgproc
linked by target "opencv_perf_cudaimgproc" in directory D:/Cproject/opencv/opencv/sources/modules/cudaimgproc
linked by target "opencv_perf_cudawarping" in directory D:/Cproject/opencv/opencv/sources/modules/cudawarping
linked by target "opencv_cudawarping" in directory D:/Cproject/opencv/opencv/sources/modules/cudawarping
linked by target "opencv_cudawarping" in directory D:/Cproject/opencv/opencv/sources/modules/cudawarping
linked by target "opencv_test_cudawarping" in directory D:/Cproject/opencv/opencv/sources/modules/cudawarping
linked by target "opencv_imgcodecs" in directory D:/Cproject/opencv/opencv/sources/modules/imgcodecs
linked by target "opencv_imgcodecs" in directory D:/Cproject/opencv/opencv/sources/modules/imgcodecs
linked by target "opencv_test_imgcodecs" in directory D:/Cproject/opencv/opencv/sources/modules/imgcodecs
linked by target "opencv_perf_imgcodecs" in directory D:/Cproject/opencv/opencv/sources/modules/imgcodecs
linked by target "opencv_perf_photo" in directory D:/Cproject/opencv/opencv/sources/modules/photo
linked by target "opencv_photo" in directory D:/Cproject/opencv/opencv/sources/modules/photo
linked by target "opencv_photo" in directory D:/Cproject/opencv/opencv/sources/modules/photo
linked by target "opencv_test_photo" in directory D:/Cproject/opencv/opencv/sources/modules/photo
linked by target "opencv_test_shape" in directory D:/Cproject/opencv/opencv/sources/modules/shape
linked by target "opencv_shape" in directory D:/Cproject/opencv/opencv/sources/modules/shape
linked by target "opencv_shape" in directory D:/Cproject/opencv/opencv/sources/modules/shape
linked by target "opencv_test_videoio" in directory D:/Cproject/opencv/opencv/sources/modules/videoio
linked by target "opencv_videoio" in directory D:/Cproject/opencv/opencv/sources/modules/videoio
linked by target "opencv_videoio" in directory D:/Cproject/opencv/opencv/sources/modules/videoio
linked by target "opencv_perf_videoio" in directory D:/Cproject/opencv/opencv/sources/modules/videoio
linked by target "opencv_cudacodec" in directory D:/Cproject/opencv/opencv/sources/modules/cudacodec
linked by target "opencv_cudacodec" in directory D:/Cproject/opencv/opencv/sources/modules/cudacodec
linked by target "opencv_perf_cudacodec" in directory D:/Cproject/opencv/opencv/sources/modules/cudacodec
linked by target "opencv_test_cudacodec" in directory D:/Cproject/opencv/opencv/sources/modules/cudacodec
linked by target "opencv_highgui" in directory D:/Cproject/opencv/opencv/sources/modules/highgui
linked by target "opencv_highgui" in directory D:/Cproject/opencv/opencv/sources/modules/highgui
linked by target "opencv_test_highgui" in directory D:/Cproject/opencv/opencv/sources/modules/highgui
linked by target "opencv_perf_objdetect" in directory D:/Cproject/opencv/opencv/sources/modules/objdetect
linked by target "opencv_objdetect" in directory D:/Cproject/opencv/opencv/sources/modules/objdetect
linked by target "opencv_objdetect" in directory D:/Cproject/opencv/opencv/sources/modules/objdetect
linked by target "opencv_test_objdetect" in directory D:/Cproject/opencv/opencv/sources/modules/objdetect
linked by target "opencv_ts" in directory D:/Cproject/opencv/opencv/sources/modules/ts
linked by target "opencv_ts" in directory D:/Cproject/opencv/opencv/sources/modules/ts
linked by target "opencv_test_features2d" in directory D:/Cproject/opencv/opencv/sources/modules/features2d
linked by target "opencv_features2d" in directory D:/Cproject/opencv/opencv/sources/modules/features2d
linked by target "opencv_features2d" in directory D:/Cproject/opencv/opencv/sources/modules/features2d
linked by target "opencv_perf_features2d" in directory D:/Cproject/opencv/opencv/sources/modules/features2d
linked by target "opencv_calib3d" in directory D:/Cproject/opencv/opencv/sources/modules/calib3d
linked by target "opencv_calib3d" in directory D:/Cproject/opencv/opencv/sources/modules/calib3d
linked by target "opencv_test_calib3d" in directory D:/Cproject/opencv/opencv/sources/modules/calib3d
linked by target "opencv_perf_calib3d" in directory D:/Cproject/opencv/opencv/sources/modules/calib3d
linked by target "opencv_cudafeatures2d" in directory D:/Cproject/opencv/opencv/sources/modules/cudafeatures2d
linked by target "opencv_cudafeatures2d" in directory D:/Cproject/opencv/opencv/sources/modules/cudafeatures2d
linked by target "opencv_test_cudafeatures2d" in directory D:/Cproject/opencv/opencv/sources/modules/cudafeatures2d
linked by target "opencv_perf_cudafeatures2d" in directory D:/Cproject/opencv/opencv/sources/modules/cudafeatures2d
linked by target "opencv_cudalegacy" in directory D:/Cproject/opencv/opencv/sources/modules/cudalegacy
linked by target "opencv_cudalegacy" in directory D:/Cproject/opencv/opencv/sources/modules/cudalegacy
linked by target "opencv_test_cudalegacy" in directory D:/Cproject/opencv/opencv/sources/modules/cudalegacy
linked by target "opencv_perf_cudalegacy" in directory D:/Cproject/opencv/opencv/sources/modules/cudalegacy
linked by target "opencv_cudaobjdetect" in directory D:/Cproject/opencv/opencv/sources/modules/cudaobjdetect
linked by target "opencv_cudaobjdetect" in directory D:/Cproject/opencv/opencv/sources/modules/cudaobjdetect
linked by target "opencv_perf_cudaobjdetect" in directory D:/Cproject/opencv/opencv/sources/modules/cudaobjdetect
linked by target "opencv_test_cudaobjdetect" in directory D:/Cproject/opencv/opencv/sources/modules/cudaobjdetect
linked by target "opencv_cudaoptflow" in directory D:/Cproject/opencv/opencv/sources/modules/cudaoptflow
linked by target "opencv_cudaoptflow" in directory D:/Cproject/opencv/opencv/sources/modules/cudaoptflow
linked by target "opencv_test_cudaoptflow" in directory D:/Cproject/opencv/opencv/sources/modules/cudaoptflow
linked by target "opencv_perf_cudaoptflow" in directory D:/Cproject/opencv/opencv/sources/modules/cudaoptflow
linked by target "opencv_perf_cudastereo" in directory D:/Cproject/opencv/opencv/sources/modules/cudastereo
linked by target "opencv_cudastereo" in directory D:/Cproject/opencv/opencv/sources/modules/cudastereo
linked by target "opencv_cudastereo" in directory D:/Cproject/opencv/opencv/sources/modules/cudastereo
linked by target "opencv_test_cudastereo" in directory D:/Cproject/opencv/opencv/sources/modules/cudastereo
linked by target "opencv_test_stitching" in directory D:/Cproject/opencv/opencv/sources/modules/stitching
linked by target "opencv_stitching" in directory D:/Cproject/opencv/opencv/sources/modules/stitching
linked by target "opencv_stitching" in directory D:/Cproject/opencv/opencv/sources/modules/stitching
linked by target "opencv_perf_stitching" in directory D:/Cproject/opencv/opencv/sources/modules/stitching
linked by target "opencv_superres" in directory D:/Cproject/opencv/opencv/sources/modules/superres
linked by target "opencv_superres" in directory D:/Cproject/opencv/opencv/sources/modules/superres
linked by target "opencv_test_superres" in directory D:/Cproject/opencv/opencv/sources/modules/superres
linked by target "opencv_perf_superres" in directory D:/Cproject/opencv/opencv/sources/modules/superres
linked by target "opencv_videostab" in directory D:/Cproject/opencv/opencv/sources/modules/videostab
linked by target "opencv_videostab" in directory D:/Cproject/opencv/opencv/sources/modules/videostab
linked by target "opencv_traincascade" in directory D:/Cproject/opencv/opencv/sources/apps/traincascade
linked by target "opencv_createsamples" in directory D:/Cproject/opencv/opencv/sources/apps/createsamples
linked by target "opencv_annotation" in directory D:/Cproject/opencv/opencv/sources/apps/annotation
linked by target "opencv_visualisation" in directory D:/Cproject/opencv/opencv/sources/apps/visualisation
linked by target "opencv_version" in directory D:/Cproject/opencv/opencv/sources/apps/version

Please set them or make sure they are set and tested correctly in the CMake files:
CUDA_nppi_LIBRARY (ADVANCED)
linked by target "opencv_cudev" in directory D:/Cproject/opencv/opencv/sources/modules/cudev
linked by target "opencv_cudev" in directory D:/Cproject/opencv/opencv/sources/modules/cudev
linked by target "opencv_test_cudev" in directory D:/Cproject/opencv/opencv/sources/modules/cudev/test
linked by target "opencv_core" in directory D:/Cproject/opencv/opencv/sources/modules/core
linked by target "opencv_core" in directory D:/Cproject/opencv/opencv/sources/modules/core
linked by target "opencv_test_core" in directory D:/Cproject/opencv/opencv/sources/modules/core
linked by target "opencv_perf_core" in directory D:/Cproject/opencv/opencv/sources/modules/core
linked by target "opencv_test_cudaarithm" in directory D:/Cproject/opencv/opencv/sources/modules/cudaarithm
linked by target "opencv_cudaarithm" in directory D:/Cproject/opencv/opencv/sources/modules/cudaarithm
linked by target "opencv_cudaarithm" in directory D:/Cproject/opencv/opencv/sources/modules/cudaarithm
linked by target "opencv_perf_cudaarithm" in directory D:/Cproject/opencv/opencv/sources/modules/cudaarithm
linked by target "opencv_flann" in directory D:/Cproject/opencv/opencv/sources/modules/flann
linked by target "opencv_flann" in directory D:/Cproject/opencv/opencv/sources/modules/flann
linked by target "opencv_test_flann" in directory D:/Cproject/opencv/opencv/sources/modules/flann
linked by target "opencv_imgproc" in directory D:/Cproject/opencv/opencv/sources/modules/imgproc
linked by target "opencv_imgproc" in directory D:/Cproject/opencv/opencv/sources/modules/imgproc
linked by target "opencv_test_imgproc" in directory D:/Cproject/opencv/opencv/sources/modules/imgproc
linked by target "opencv_perf_imgproc" in directory D:/Cproject/opencv/opencv/sources/modules/imgproc
linked by target "opencv_test_ml" in directory D:/Cproject/opencv/opencv/sources/modules/ml
linked by target "opencv_ml" in directory D:/Cproject/opencv/opencv/sources/modules/ml
linked by target "opencv_ml" in directory D:/Cproject/opencv/opencv/sources/modules/ml
linked by target "opencv_test_video" in directory D:/Cproject/opencv/opencv/sources/modules/video
linked by target "opencv_video" in directory D:/Cproject/opencv/opencv/sources/modules/video
linked by target "opencv_video" in directory D:/Cproject/opencv/opencv/sources/modules/video
linked by target "opencv_perf_video" in directory D:/Cproject/opencv/opencv/sources/modules/video
linked by target "opencv_cudabgsegm" in directory D:/Cproject/opencv/opencv/sources/modules/cudabgsegm
linked by target "opencv_cudabgsegm" in directory D:/Cproject/opencv/opencv/sources/modules/cudabgsegm
linked by target "opencv_test_cudabgsegm" in directory D:/Cproject/opencv/opencv/sources/modules/cudabgsegm
linked by target "opencv_perf_cudabgsegm" in directory D:/Cproject/opencv/opencv/sources/modules/cudabgsegm
linked by target "opencv_cudafilters" in directory D:/Cproject/opencv/opencv/sources/modules/cudafilters
linked by target "opencv_cudafilters" in directory D:/Cproject/opencv/opencv/sources/modules/cudafilters
linked by target "opencv_test_cudafilters" in directory D:/Cproject/opencv/opencv/sources/modules/cudafilters
linked by target "opencv_perf_cudafilters" in directory D:/Cproject/opencv/opencv/sources/modules/cudafilters
linked by target "opencv_cudaimgproc" in directory D:/Cproject/opencv/opencv/sources/modules/cudaimgproc
linked by target "opencv_cudaimgproc" in directory D:/Cproject/opencv/opencv/sources/modules/cudaimgproc
linked by target "opencv_test_cudaimgproc" in directory D:/Cproject/opencv/opencv/sources/modules/cudaimgproc
linked by target "opencv_perf_cudaimgproc" in directory D:/Cproject/opencv/opencv/sources/modules/cudaimgproc
linked by target "opencv_perf_cudawarping" in directory D:/Cproject/opencv/opencv/sources/modules/cudawarping
linked by target "opencv_cudawarping" in directory D:/Cproject/opencv/opencv/sources/modules/cudawarping
linked by target "opencv_cudawarping" in directory D:/Cproject/opencv/opencv/sources/modules/cudawarping
linked by target "opencv_test_cudawarping" in directory D:/Cproject/opencv/opencv/sources/modules/cudawarping
linked by target "opencv_imgcodecs" in directory D:/Cproject/opencv/opencv/sources/modules/imgcodecs
linked by target "opencv_imgcodecs" in directory D:/Cproject/opencv/opencv/sources/modules/imgcodecs
linked by target "opencv_test_imgcodecs" in directory D:/Cproject/opencv/opencv/sources/modules/imgcodecs
linked by target "opencv_perf_imgcodecs" in directory D:/Cproject/opencv/opencv/sources/modules/imgcodecs
linked by target "opencv_perf_photo" in directory D:/Cproject/opencv/opencv/sources/modules/photo
linked by target "opencv_photo" in directory D:/Cproject/opencv/opencv/sources/modules/photo
linked by target "opencv_photo" in directory D:/Cproject/opencv/opencv/sources/modules/photo
linked by target "opencv_test_photo" in directory D:/Cproject/opencv/opencv/sources/modules/photo
linked by target "opencv_test_shape" in directory D:/Cproject/opencv/opencv/sources/modules/shape
linked by target "opencv_shape" in directory D:/Cproject/opencv/opencv/sources/modules/shape
linked by target "opencv_shape" in directory D:/Cproject/opencv/opencv/sources/modules/shape
linked by target "opencv_test_videoio" in directory D:/Cproject/opencv/opencv/sources/modules/videoio
linked by target "opencv_videoio" in directory D:/Cproject/opencv/opencv/sources/modules/videoio
linked by target "opencv_videoio" in directory D:/Cproject/opencv/opencv/sources/modules/videoio
linked by target "opencv_perf_videoio" in directory D:/Cproject/opencv/opencv/sources/modules/videoio
linked by target "opencv_cudacodec" in directory D:/Cproject/opencv/opencv/sources/modules/cudacodec
linked by target "opencv_cudacodec" in directory D:/Cproject/opencv/opencv/sources/modules/cudacodec
linked by target "opencv_perf_cudacodec" in directory D:/Cproject/opencv/opencv/sources/modules/cudacodec
linked by target "opencv_test_cudacodec" in directory D:/Cproject/opencv/opencv/sources/modules/cudacodec
linked by target "opencv_highgui" in directory D:/Cproject/opencv/opencv/sources/modules/highgui
linked by target "opencv_highgui" in directory D:/Cproject/opencv/opencv/sources/modules/highgui
linked by target "opencv_test_highgui" in directory D:/Cproject/opencv/opencv/sources/modules/highgui
linked by target "opencv_perf_objdetect" in directory D:/Cproject/opencv/opencv/sources/modules/objdetect
linked by target "opencv_objdetect" in directory D:/Cproject/opencv/opencv/sources/modules/objdetect
linked by target "opencv_objdetect" in directory D:/Cproject/opencv/opencv/sources/modules/objdetect
linked by target "opencv_test_objdetect" in directory D:/Cproject/opencv/opencv/sources/modules/objdetect
linked by target "opencv_ts" in directory D:/Cproject/opencv/opencv/sources/modules/ts
linked by target "opencv_ts" in directory D:/Cproject/opencv/opencv/sources/modules/ts
linked by target "opencv_test_features2d" in directory D:/Cproject/opencv/opencv/sources/modules/features2d
linked by target "opencv_features2d" in directory D:/Cproject/opencv/opencv/sources/modules/features2d
linked by target "opencv_features2d" in directory D:/Cproject/opencv/opencv/sources/modules/features2d
linked by target "opencv_perf_features2d" in directory D:/Cproject/opencv/opencv/sources/modules/features2d
linked by target "opencv_calib3d" in directory D:/Cproject/opencv/opencv/sources/modules/calib3d
linked by target "opencv_calib3d" in directory D:/Cproject/opencv/opencv/sources/modules/calib3d
linked by target "opencv_test_calib3d" in directory D:/Cproject/opencv/opencv/sources/modules/calib3d
linked by target "opencv_perf_calib3d" in directory D:/Cproject/opencv/opencv/sources/modules/calib3d
linked by target "opencv_cudafeatures2d" in directory D:/Cproject/opencv/opencv/sources/modules/cudafeatures2d
linked by target "opencv_cudafeatures2d" in directory D:/Cproject/opencv/opencv/sources/modules/cudafeatures2d
linked by target "opencv_test_cudafeatures2d" in directory D:/Cproject/opencv/opencv/sources/modules/cudafeatures2d
linked by target "opencv_perf_cudafeatures2d" in directory D:/Cproject/opencv/opencv/sources/modules/cudafeatures2d
linked by target "opencv_cudalegacy" in directory D:/Cproject/opencv/opencv/sources/modules/cudalegacy
linked by target "opencv_cudalegacy" in directory D:/Cproject/opencv/opencv/sources/modules/cudalegacy
linked by target "opencv_test_cudalegacy" in directory D:/Cproject/opencv/opencv/sources/modules/cudalegacy
linked by target "opencv_perf_cudalegacy" in directory D:/Cproject/opencv/opencv/sources/modules/cudalegacy
linked by target "opencv_cudaobjdetect" in directory D:/Cproject/opencv/opencv/sources/modules/cudaobjdetect
linked by target "opencv_cudaobjdetect" in directory D:/Cproject/opencv/opencv/sources/modules/cudaobjdetect
linked by target "opencv_perf_cudaobjdetect" in directory D:/Cproject/opencv/opencv/sources/modules/cudaobjdetect
linked by target "opencv_test_cudaobjdetect" in directory D:/Cproject/opencv/opencv/sources/modules/cudaobjdetect
linked by target "opencv_cudaoptflow" in directory D:/Cproject/opencv/opencv/sources/modules/cudaoptflow
linked by target "opencv_cudaoptflow" in directory D:/Cproject/opencv/opencv/sources/modules/cudaoptflow
linked by target "opencv_test_cudaoptflow" in directory D:/Cproject/opencv/opencv/sources/modules/cudaoptflow
linked by target "opencv_perf_cudaoptflow" in directory D:/Cproject/opencv/opencv/sources/modules/cudaoptflow
linked by target "opencv_perf_cudastereo" in directory D:/Cproject/opencv/opencv/sources/modules/cudastereo
linked by target "opencv_cudastereo" in directory D:/Cproject/opencv/opencv/sources/modules/cudastereo
linked by target "opencv_cudastereo" in directory D:/Cproject/opencv/opencv/sources/modules/cudastereo
linked by target "opencv_test_cudastereo" in directory D:/Cproject/opencv/opencv/sources/modules/cudastereo
linked by target "opencv_test_stitching" in directory D:/Cproject/opencv/opencv/sources/modules/stitching
linked by target "opencv_stitching" in directory D:/Cproject/opencv/opencv/sources/modules/stitching
linked by target "opencv_stitching" in directory D:/Cproject/opencv/opencv/sources/modules/stitching
linked by target "opencv_perf_stitching" in directory D:/Cproject/opencv/opencv/sources/modules/stitching
linked by target "opencv_superres" in directory D:/Cproject/opencv/opencv/sources/modules/superres
linked by target "opencv_superres" in directory D:/Cproject/opencv/opencv/sources/modules/superres
linked by target "opencv_test_superres" in directory D:/Cproject/opencv/opencv/sources/modules/superres
linked by target "opencv_perf_superres" in directory D:/Cproject/opencv/opencv/sources/modules/superres
linked by target "opencv_videostab" in directory D:/Cproject/opencv/opencv/sources/modules/videostab
linked by target "opencv_videostab" in directory D:/Cproject/opencv/opencv/sources/modules/videostab
linked by target "opencv_traincascade" in directory D:/Cproject/opencv/opencv/sources/apps/traincascade
linked by target "opencv_createsamples" in directory D:/Cproject/opencv/opencv/sources/apps/createsamples
linked by target "opencv_annotation" in directory D:/Cproject/opencv/opencv/sources/apps/annotation
linked by target "opencv_visualisation" in directory D:/Cproject/opencv/opencv/sources/apps/visualisation
linked by target "opencv_version" in directory D:/Cproject/opencv/opencv/sources/apps/version

参考网址

https://blog.csdn.net/u014613745/article/details/78310916

3.20 git clone速度太慢解决方案

参考网址

https://blog.csdn.net/hzwwpgmwy/article/details/79043251

3.21 ImportError:No module named skimage.io

参考网址

https://blog.csdn.net/m0_37407756/article/details/70789271 （错误5）

3.22 error while loading shared libraries: libcudnn.so.7: cannot open shared object file: No such file or

参考网址

https://blog.csdn.net/dc1994dc/article/details/79162886

https://blog.csdn.net/sinat_23619409/article/details/85047788

3.23 import caffe失败 No module named caffe

参考网址

https://blog.csdn.net/a8039974/article/details/79708457

3.24 caffe2安装教程的借鉴原文

参考网址

https://blog.csdn.net/lemonaha/article/details/71436327