OpenCV on Linux using g++, CMake, Qt, Code::Blocks

1. Introduction
2. Using GCC ၊ CMake
3. Using Qt
4. Using Code::Blocks
5. References

Introduction

To install OpenCV on your Linux machine, you need to have the following packages as prerequisites [Ope17g].

• GCC 4.4.x or later
• CMake 2.6 or higher
• Git
• GTK+2.x or higher, including headers (libgtk2.0-dev)
• pkg-config
• Python 2.6 or later and Numpy 1.5 or later with developer packages (python-dev, python-numpy)
• ffmpeg or libav development packages: libavcodec-dev, libavformat-dev, libswscale-dev
• [optional] libtbb2 libtbb-dev
• [optional] libdc1394 2.x
• [optional] libjpeg-dev, libpng-dev, libtiff-dev, libjasper-dev, libdc1394-22-dev

To install the required packages, you can enter the following commands in terminal.

sudo apt-get update
sudo apt-get install build-essential
sudo apt-get install cmake git libgtk-3-dev pkg-config
sudo apt-get install libavcodec-dev libavformat-dev libswscale-dev

Listing. Getting required packages for OpenCV.

sudo apt-get install python-dev python-numpy libtbb2 libtbb-dev libjpeg-dev
sudo apt-get install libpng-dev libtiff-dev libdc1394-22-dev
sudo apt install libv4l-dev libxvidcore-dev libx264-dev

Listing. Getting optional packages for OpenCV.

As a simple way to install OpenCV on Linux, you can enter the following command in terminal.

sudo apt-get install libopencv-dev

If not, you can get the current stable version OpenCV at OpenCV for Linux/Mac (https://opencv.org/). Then, you can use Archive Manager to extract the zip file.

If you want the latest cutting-edge opencv version, you can get it from Git repository OpenCV repository by using following commands instead of using above methods. OpenCV contrib repository can also be obtained as follows.

cd ~
git clone https://github.com/opencv/opencv.git
git clone https://github.com/opencv/opencv_contrib.git

Navigate to the resulting opencv directory and make a build directory.

cd ~/opencv
mkdir build
cd build
cmake -D CMAKE_BUILD_TYPE=Release \
-D CMAKE_INSTALL_PREFIX=/usr/local \
-D OPENCV_EXTRA_MODULES_PATH=~/opencv_contrib/modules ..

Listing. Building OpenCV.

You can use the following option if you want static libs.

-D BUILD_SHARED_LIBS=OFF

Listing. Option to use static libs.

After a while, you can build and install it as follows.

make
sudo make install

Listing. Installing OpenCV

Check the opencv version as follow.

pkg-config --modversion opencv

Listing. OpenCV version

Using GCC ၊ CMake

You can use CMake to build your OpenCV applications [Ope17l]. You might want to have a look at CMake tutorial (https://cmake.org/cmake-tutorial/) to learn cmake.

As a simple example, DisplayImage.cpp (online) reads an image and displays it. For that, an image file, thiri.jpg, is placed in home folder. A folder can be created to put DisplayImage.cpp.

#include < stdio.h >
#include < opencv2/opencv.hpp >
using namespace cv;
int main(int argc, char** argv )
{
 Mat image;
 image = imread( "/home/yan/thiri.jpg", 1 );
 if ( !image.data ) {
  printf("No image data \n");
  return -1;
 }
 namedWindow("Display Image", WINDOW_AUTOSIZE );
 imshow("Display Image", image);
 waitKey(0);
 return 0;
}

Listing. DisplayImage.cpp

In this example, the path for thiri.jpg is "/home/yan/thiri.jpg". You can replace it with your own image path. The function 'imread' loads the image. The value 1 for second argument is to read the image in color. The value can be 0 to read image in grayscale. If the reading of the image is failed, the program will exit with an error message. If it is successful, the image will be displayed using imshow. Create a file called CMakeLists.txt in the same folder as follow.

cmake_minimum_required(VERSION 2.8)
project( DisplayImage )
find_package( OpenCV REQUIRED )
add_executable( DisplayImage DisplayImage.cpp )
target_link_libraries( DisplayImage ${OpenCV_LIBS} )

Listing. CMakeLists.txt

Enter the following commands in terminal to build and run the program.

cd DisplayImage
cmake .
make
./DisplayImage

Listing. Building DisplayImage with CMake, and running it.

pkg-config

You can also use 'pkg-config' instead of CMake to build and run the application as follow.

g++ DisplayImage.cpp `pkg-config --cflags --libs opencv` -o DisplayImage
./DisplayImage

Listing. Using g++, pkg-config to build and run the application

The following Figure illustrates the output the program.

Figure. The output of DisplayImage program.

If you get an error of missing shared lib, /etc/ld.so.conf.d/opencv.conf can be created as follow.

sudo nano /etc/ld.so.conf.d/opencv.conf

Then, input the opencv path depending on your system.

/usr/local/opencv/

or

/usr/local/lib/

Enter the following command to load the configuration.

sudo ldconfig

Using Qt

Let us discuss about using OpenCV with Qt. The following commands can be used to set up Qt.

wget http://download.qt.io/official_releases/qt/5.9/5.9.0/qt-opensource-linux-x64-5.9.0.run
chmod +x qt-opensource-linux-x64-5.9.0.run
./qt-opensource-linux-x64-5.9.0.run

Listing. Installing Qt on Linux.

Other required packages such as g++, font configuration, and OpenGL can be acquired as follows.

sudo apt-get install build-essential
sudo apt-get install libfontconfig1
sudo apt-get install mesa-common-dev
sudo apt-get install libglu1-mesa-dev -y

Listing. Getting required packages for Qt.

Open Qt to test a simple OpenCV example. Click New Project, and choose Non-Qt Project and Plain C++ Application as shown in the following Figure.

Figure. Creating a new project in Qt.

For the 'Project Location', give a name like DisplayImageQt and create a folder. Leave qmake for Build System and Desktop Qt 5.9.0 GCC 64bit for Kit Selection. Click Next and leave None for Project Management. Then, write a program in Project → Sources → main.cpp (online) as follow.

#include < stdio.h >
#include < opencv2/opencv.hpp >
using namespace cv;
int main(int argc, char** argv )
{
 Mat image,bw_img;
 image = imread( "/home/yan/thiri.jpg", 1 );
 if ( !image.data ) {
  printf("No image data \n");
  return -1;
 }
 namedWindow("Original Image", WINDOW_AUTOSIZE );
 imshow("Original Image", image);
 cvtColor(image, bw_img, COLOR_BGR2GRAY);
 namedWindow("BW Image", WINDOW_AUTOSIZE );
 imshow("BW Image", bw_img);
 waitKey(0);
 return 0;
}

Listing. A program to read an image and convert it to grayscale.

The program read an image and convert it to grayscale using 'cvtColor'. To build the project, open DisplayImageQt.pro and modify it as follows.

TEMPLATE = app
CONFIG += console c++11
CONFIG -= app_bundle
CONFIG -= qt
SOURCES += main.cpp
INCLUDEPATH += /usr/local/include/opencv
LIBS += -L/usr/local/lib -lopencv_core -lopencv_highgui
 -lopencv_imgproc -lopencv_imgcodecs

Listing. DisplayImageQt.pro

When the program is run, a grayscale image of Thiri will appear beautifully.

Figure. The output of the program showing original and converted images of Thiri.

Using Code::Blocks

Code::Blocks is a popular IDE for C++. In this section, using Code::Blocks with OpenCV is discussed. To install Code::Blocks IDE, enter the following commands in terminal.

sudo apt-get install build-essential
sudo apt-get install libgtk-3-dev
sudo apt-get install gdb
sudo add-apt-repository ppa:damien-moore/codeblocks-stable
sudo apt-get update
sudo apt-get install codeblocks codeblocks-contrib

Listing. Installing Code::Blocks.

After installing Code::Blocks, open the IDE and create a new project by clicking 'File menu → New → Project...' or 'Start here tab → Create a new project'. A pop up window will appear and choose 'OpenCV project' for Category.

Figure. Creating an OpenCV project in Code::Blocks IDE.

Click next and define project name and location.

Figure. Defining project name and location.

Then, choose compiler as 'GNU GCC compiler' as shown and click Finish.

Figure. Choosing Compiler.

Thereafter, you can see OpenCV sample code and lena.jpg in the project. Modify the header as #include < opencv2/opencv.hpp > in main.cpp as shown in the following Figure and Listing (online).

Figure. Modifying the sample code.

#include < opencv2/opencv.hpp >
using namespace cv;
int main(int argc, char *argv[])
{
    Mat img = imread("lena.jpg", CV_LOAD_IMAGE_COLOR);
    if(img.empty())
       return -1;
    namedWindow( "lena", CV_WINDOW_AUTOSIZE );
    imshow("lena", img);
    waitKey(0);
    return 0;
}

Listing. An example OpenCV program for Code::Blocks.

Defining Build Options

Click Project Menu → Build Options... and define /usr/local/include in Search directories tab → Compiler tab for both Debug and Release configurations as illustrated in the following Figure.

Figure. Defining Compiler Search directories.

Similarly, define /usr/local/lib in Linker tab as in the following Figure.

Figure. Defining Search directories for Libraries.

As a next step, enter Link Libraries in Linker settings tab as illustrated in the following Figure.

Figure. Adding Link libraries.

After all these steps, click F9 to build and run the program. Its output can be seen as shown in the following Figure.

Figure. Running the example OpenCV program in Code::Blocks.

Using pkg-config

You can also use pkg-config to define build options. Click Project Menu → Build Options... and enter the following line in Compiler settings tab → Other compiler options tab for both Debug and Release configurations.

`pkg-config --cflags opencv`

Figure. Using pkg-config for compiler settings.

Similarly, enter the following line in Linker settings tab → Other linker options tab.

`pkg-config --libs opencv`

Figure. Using pkg-config for linker settings.

When you build and run the project (online) by clicking F9 , the same output will appear as shown in the previous method.

Related Posts

• OpenCV on Windows Using Visual Studio
• Real-time Face Detection with OpenCV

References

[Ope17g] OpenCV. Installation in Linux. 2017.
url: http://docs.opencv.org/2.4/doc/tutorials/introduction/linux_install/linux_install.html.

[Ope17l] OpenCV. Using OpenCV with gcc and CMake. 2017.
url: http://docs.opencv.org/2.4/doc/tutorials/introduction/linux_gcc_cmake/linux_gcc_cmake.html.