Ubuntu14+OpenCV3.1+Caffe虚拟机安装笔记（二）

2. 安装OpenCV3.1

OpenCV参考资料

 

OpenCV安装手册：http://www.learnopencv.com/install-opencv3-on-ubuntu/

OpenCV安装参考： http://blog.csdn.net/yangdashi888/article/details/52088264

Learn OpenCV : C++ and Python Examples

https://github.com/spmallick/learnopencv

 

 

Step 1: Update packages

sudo apt-get update

sudo apt-get upgrade

 

Step 2: Install OS libraries

 

Step 2.1: 删除以前安装的任何 x264

sudo apt-get remove x264 libx264-dev

 

Step 2.2: Install dependencies

sudo apt-get install build-essential checkinstall cmakepkg-config yasm gfortran git

sudo apt-get install libjpeg8-dev libjasper-dev libpng12-dev

sudo apt-get install libtiff4-dev

sudo apt-get install libavcodec-dev libavformat-devlibswscale-dev libdc1394-22-dev

sudo apt-get install libxine2-dev libv4l-dev

sudo apt-get install libgstreamer0.10-devlibgstreamer-plugins-base0.10-dev

sudo apt-get install libqt4-dev libgtk2.0-dev libtbb-dev

sudo apt-get install libatlas-base-dev

sudo apt-get install libfaac-dev libmp3lame-devlibtheora-dev

sudo apt-get install libvorbis-dev libxvidcore-dev

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

sudo apt-get install x264 v4l-utils

 

Step 2.3: Optional dependencies

sudo apt-get install libprotobuf-dev protobuf-compiler

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

sudo apt-get install libgphoto2-dev libeigen3-devlibhdf5-dev doxygen

 

Step2.4给用户moses设定root权限

su - root

gedit /etc/sudoers

加入：

# User privilege specification

root       ALL=(ALL:ALL)ALL

moses   ALL=(ALL:ALL)ALL

 

Step2.5安装ffmpeg

sudoadd-apt-repository ppa:mc3man/trusty-media[s1] 

$ sudo apt-get install ffmpeg gstreamer0.10-ffmpeg

$ sudo apt-get install libx264-dev

 

Step2.6安装Java、ant

1. 下载jdk

http://www.Oracle.com/technetwork/Java/javase/downloads/jdk8-downloads-2133151.html

并把下载文件解压缩到/usr/local/jdk1.8.0_131

apache-ant-1.10.1

 

2. sudo gedit /etc/profile,最后加入：

export ANT_HOME=/usr/local/apache-ant

export JAVA_HOME=/usr/local/jdk1.8.0_131

export JRE_HOME=$JAVA_HOME/jre

export PATH=$JAVA_HOME/bin:$JRE_HOME/bin:$PATH:$ANT_HOME/bin

export CLASSPATH=.:$JAVA_HOME/lib:$JRE_HOME/lib

 

3. 使配置生效 source /etc/profile

 

 

 

Step 3: Install Python libraries

 

升级Python：

Ubuntu14.04默认的python2.7是Python2.7.6，python3是Python 3.4.3，需要升级才能进行后面的安装

$ sudo add-apt-repository ppa:fkrull/deadsnakes-python2.7

$ sudo apt-get update  

$ sudo apt-get upgrade

 

sudo add-apt-repository ppa:fkrull/deadsnakes

$sudo apt-get update

$sudo apt-get upgrade python3.5          //也可以装3.6，看个人喜好了。

 

 

sudo apt-get install python-dev python-pip python3-devpython3-pip

sudo-H pip2 install -U pip numpy

sudo -H pip3 install -U pip numpy

如果出现错误：

Cannot fetch index base URL https://pypi.python.org/simple/

Could not find any downloads that satisfy the requirementpip in /usr/lib/python2.7/dist-packages，

解决：

新建文件~/.pip/pip.conf，并添加以下内容。

[global]

index-url = http://pypi.douban.com/simple

[install]

trusted-host = pypi.douban.com

 

We will use Virtual Environment to install Python libraries.It is generally a good practice in order to separate your project environmentand global environment.

 

# Install virtual environment

sudo pip2 install virtualenvvirtualenvwrapper

第二次安装的时候，报错：The directory '/home/moses/.cache/pip/http'or its parent directory is not owned by the current user and the cache has beendisabled. Please check the permissions and owner of that directory. Ifexecuting pip with sudo, you may want sudo's -H flag.

解决：sudo -H pip2install virtualenv virtualenvwrapper

 

sudo pip3 install virtualenv virtualenvwrapper

echo "# Virtual Environment Wrapper"  >> ~/.bashrc

echo "source /usr/local/bin/virtualenvwrapper.sh">> ~/.bashrc

source ~/.bashrc

 

############ For Python 2 ############

# create virtual environment

mkvirtualenvfacecourse-py2 -p python2[s6] 

workon facecourse-py2

 

# now install python libraries within this virtualenvironment

pipinstall numpy scipy matplotlib scikit-image scikit-learn ipython

报错1：

Exception:

Traceback(most recent call last):

  File "/home/moses/.virtualenvs/facecourse-py2/local/lib/python2.7/site-packages/pip/basecommand.py",line 215, in main

 

报错2：ImportError: No module named urllib3

 (facecourse-py2)moses@ubuntu:/$ sudo pip install --upgrade requests

 

报错3：Command"python setup.py egg_info" failed with error code 1 in/tmp/pip-build-0U6ZZT/ipython/

解决：$ pipinstall --upgrade setuptools

 

报错4：IPython 6.0+ does not support Python 2.6,2.7, 3.0, 3.1, or 3.2.

    When using Python 2.7, please installIPython 5.x LTS Long Term Support version.

    Beginning with IPython 6.0, Python 3.3 andabove is required.

解决：

(facecourse-py2) moses@ubuntu:/$ pip install numpy scipymatplotlib scikit-image scikit-learn

(facecourse-py2) moses@ubuntu:/$ sudo apt-get installipython

 

# quit virtual environment

(facecourse-py2) moses@ubuntu:/$ deactivate

######################################

 

############ For Python 3 ############

# create virtual environment

mkvirtualenv facecourse-py3 -p python3

workon facecourse-py3

# now install python libraries within this virtualenvironment

pip install numpy scipy matplotlib scikit-image scikit-learnipython

 

# quit virtual environment

deactivate

######################################

 

 

 

 

Step 4: Download OpenCV and OpenCV_contrib

 

We will download opencv and opencv_contribpackages from their GitHub repositories.

 

Step 4.1: Download opencv from Github

 

If you have OpenBlas installed on yourmachine, OpenCV 3.2.0 fails to compile. We will use the commit where this bughas been patched. So this is OpenCV 3.2.0 with few bugs patched.

 

git clonehttps://github.com/opencv/opencv.git

cd opencv

git checkout2b44c0b6493726c465152e1db82cd8e65944d0db

cd ..

Step 4.2: Download opencv_contrib fromGithub

 

If we use v3.2.0, python module ofopencv (cv2) fails to import due to a bug in opencv_contrib. Here too we willuse a commit where this bug has been patched.

 

git clonehttps://github.com/opencv/opencv_contrib.git

cd opencv_contrib

git checkoutabf44fcccfe2f281b7442dac243e37b7f436d961

cd ..[s8] 

 

 

Step 5: Compile and install OpenCV with contrib modules

 

Step 5.1: Create a build directory inside OpenCV directory

 

cd opencv

mkdir build

cd build

 

Step 5.2: Run CMake

 

$ sudo apt-get install cmake-qt-gui             #cmake图形化界面

$ cmake-gui                                                     #运行图形化cmake

 

cmake-D CMAKE_BUILD_TYPE=RELEASE \

      -D CMAKE_INSTALL_PREFIX=/usr/local \

      -D INSTALL_C_EXAMPLES=ON \

      -D INSTALL_PYTHON_EXAMPLES=ON \

      -D WITH_TBB=ON \

      -D WITH_V4L=ON \

      -D WITH_QT=ON \

      -D WITH_OPENGL=ON \

      -DOPENCV_EXTRA_MODULES_PATH=../../opencv_contrib/modules \

      -D BUILD_EXAMPLES=ON ..

第一次make会报错，因为下载不到ippicv_linux_20151201.tgz，第二次编译前，把文件ippicv_linux_20151201.tgz，复制到/opt/opencv/3rdparty/ippicv/downloads/linux-808b791a6eac9ed78d32a7666804320e目录下。

 

升级cmake

$ sudo add-apt-repository ppa:george-edison55/cmake-3.x

$ sudo apt-get update

$ sudo apt-get upgrade

$ sudo apt-get --reinstall install cmake

 

 

 

 

Step 5.3: Compile and Install

 

# find out number of CPU cores in your machine

nproc

# substitute 4 by output of nproc

make-j4

参考资料：http://blog.csdn.net/eagelangel/article/details/51530425

文件：/opt/opencv/modules/highgui/src/window_QT.cpp中，mouse_wheel和QPoingF没有定义

参考资料：https://github.com/opencv/opencv/issues/6993

我的解决方案：在cmake里面把QT给关了……郁闷啊

cmake -D CMAKE_BUILD_TYPE=RELEASE \

      -D CMAKE_INSTALL_PREFIX=/usr/local \

      -D INSTALL_C_EXAMPLES=ON \

      -D INSTALL_PYTHON_EXAMPLES=ON \

      -D WITH_TBB=ON \

      -D WITH_V4L=ON \

      -D WITH_QT=OFF \

      -D WITH_OPENGL=ON \

      -DOPENCV_EXTRA_MODULES_PATH=../../opencv_contrib/modules \

      -D BUILD_EXAMPLES=ON ..

 

 

 

$ sudo make install

$ sudo sh -c 'echo "/usr/local/lib" >>/etc/ld.so.conf.d/opencv.conf'

$ sudo ldconfig

 

Step 5.4: Create symlink in virtual environment

 

Depending upon Python version you have, paths would bedifferent. OpenCV’s Python binary (cv2.so) can be installed either in directorysite-packages or dist-packages. Use the following command to find out thecorrect location on your machine.

 

$ find/usr/local/lib/ -type f -name "cv2*.so"

/usr/local/lib/python3.4/dist-packages/cv2.cpython-34m.so

/usr/local/lib/python2.7/dist-packages/cv2.so

 

It shouldoutput paths similar to one of these (or two in case OpenCV was compiled forboth Python2 and Python3:

############For Python 2 ############

## binaryinstalled in dist-packages

/usr/local/lib/python2.7/dist-packages/cv2.so

## binaryinstalled in site-packages

/usr/local/lib/python2.7/site-packages/cv2.so

 

############For Python 3 ############

## binaryinstalled in dist-packages

/usr/local/lib/python3.5/dist-packages/cv2.cpython-35m-x86_64-linux-gnu.so

/usr/local/lib/python3.6/dist-packages/cv2.cpython-36m-x86_64-linux-gnu.so

## binaryinstalled in site-packages

/usr/local/lib/python3.5/site-packages/cv2.cpython-35m-x86_64-linux-gnu.so

/usr/local/lib/python3.6/site-packages/cv2.cpython-36m-x86_64-linux-gnu.so

Double checkthe exact path on your machine before running the following commands

 

 

############ For Python 2 ############

cd ~/.virtualenvs/facecourse-py2/lib/python2.7/site-packages

ln -s /usr/local/lib/python2.7/dist-packages/cv2.so cv2.so

 

############ For Python 3 ############

cd ~/.virtualenvs/facecourse-py3/lib/python3.4/site-packages

ln -s /usr/local/lib/python3.4/dist-packages/cv2.cpython-36m-x86_64-linux-gnu.socv2.so

 

Step 6: Test OpenCV3.1

 

We will test a red eye remover application written in OpenCVto test our C++ and Python installations. Download RedEyeRemover.zip andextract it into a folder.

 

Step 6.1: Test C++ code

 

Move inside extracted folder, compile and run.

 

# compile

# There are backticks ( ` ) around pkg-config command not singlequotes

g++ -std=c++11 removeRedEyes.cpp `pkg-config --libs --cflagsopencv` -o removeRedEyes

# run

./removeRedEyes

参考资料：http://www.learnopencv.com/automatic-red-eye-remover-using-opencv-cpp-python/

 

Step 6.2: Test Python code

 

Activate Python virtual environment

############ For Python 2 ############

workon facecourse-py2

 

############ For Python 3 ############

workon facecourse-py3

Quick Check

 

# open ipython (run this command on terminal)

ipython

# import cv2 and print version (run following commands inipython)

import cv2

print cv2.__version__

# If OpenCV3 is installed correctly, the above commandshould give output 3.2.0

# Press CTRL+D to exit ipython

Run RedEyeRemover demo

 

python removeRedEyes.py

Now you can exit from Python virtual environment.

 

deactivate

Whenever you are going to run Python scripts which useOpenCV you should activate the virtual environment we created, using workoncommand.

 

---

[ 38%] Built target opencv_superres

/opt/opencv/modules/highgui/src/window_QT.cpp: Inmember function ‘virtual void DefaultViewPort::wheelEvent(QWheelEvent*)’:

/opt/opencv/modules/highgui/src/window_QT.cpp:2637:27:error: ‘mouse_wheel’ was not declared in this scope

     icvmouseHandler(evnt, mouse_wheel, cv_event,flags);

                           ^

/opt/opencv/modules/highgui/src/window_QT.cpp:2638:34:error: ‘QPoingF’ was not declared in this scope

    icvmouseProcessing(QPoingF(pt), cv_event, flags);

                                  ^

/opt/opencv/modules/highgui/src/window_QT.cpp:2643:5:error: ‘mouseWheel’ is not a member of ‘QWidget’

    QWidget::mouseWheel(evnt);

     ^

make[2]: ***[modules/highgui/CMakeFiles/opencv_highgui.dir/src/window_QT.cpp.o] Error 1

make[2]: *** Waiting for unfinished jobs....

make[1]: ***[modules/highgui/CMakeFiles/opencv_highgui.dir/all] Error 2

make: *** [all] Error 2