[视觉算法] 妙算上cuda、opencv以及qt的配置

一、安装CUDA(同官方PDF)
1. 访问下链接:
http://developer.download.nvidia ... uda-repo-l4t-r21.3-
6-5-prod_6.5-42_armhf.deb
2. 执行以下命令安装 “cuda-repo-l4t-r21.3-6-5-prod_6.5-42_armhf.deb”至操作系统。

[plain] view plain copy

                               
登录/注册后可看大图

                               
登录/注册后可看大图

• sudo dpkg -i cuda-repo-l4t-r21.3-6-5-prod_6.5-42_armhf.deb  
• sudo apt-get update  
• sudo apt-get install cuda-toolkit-6-5  
  

具体请参考以下链接:https://developer.nvidia.com/embedded/linux-tegra

（想跑CUDA例程的可参考http://blog.csdn.net/FRD2009041510/article/details/42042807）

二、安装OpenCV

（这一步主要参照两个网站

1.http://blog.csdn.net/FRD2009041510/article/details/42930113   //图文教程挺详细的

2.http://elinux.org/Jetson/Installing_OpenCV）

1.安装 OpenCV4tegra
访问以下链接以下载安装包“ libopencv4tegra-repo_l4t-r21_2.4.10.1_armhf.
deb ”
http://developer.download.nvidia ... bopencv4tegra-repo_
l4t-r21_2.4.10.1_armhf.deb   //最新的库版本是2.4.12了
执行以下命令进行安装

[plain] view plain copy

                               
登录/注册后可看大图

                               
登录/注册后可看大图

• sudo dpkg -i libopencv4tegra-repo_l4t-r21_2.4.10.1_armhf.deb  
• sudo apt-get update  
• sudo apt-get install libopencv4tegra libopencv4tegra-dev  
  

2.安装函数库

[plain] view plain copy

                               
登录/注册后可看大图

                               
登录/注册后可看大图

• sudo add-apt-repository universe  
• sudo apt-get update  
• # Some general development libraries  
• sudo apt-get install build-essential make cmake cmake-curses-gui g++  
• # libav video input/output development libraries  
• sudo apt-get install libavformat-dev libavutil-dev libswscale-dev  
• # Video4Linux camera development libraries  
• sudo apt-get install libv4l-dev  
• # Eigen3 math development libraries  
• sudo apt-get install libeigen3-dev  
• # OpenGL development libraries (to allow creating graphical windows)  
• sudo apt-get install libglew1.6-dev  
• # GTK development libraries (to allow creating graphical windows)  
• sudo apt-get install libgtk2.0-dev  
  

3.安装opencv源码

①.下载源码（推荐2.4.9）

http://downloads.sourceforge.net/project/opencvlibrary/opencv-unix/2.4.9/opencv-2.4.9.zip

②.解压

[plain] view plain copy

                               
登录/注册后可看大图

                               
登录/注册后可看大图

• cd Downloads  
• unzip opencv-2.4.9.zip  
• mv opencv-2.4.9 ~  
  

③.Make

[plain] view plain copy

                               
登录/注册后可看大图

                               
登录/注册后可看大图

• cd ~/opencv-2.4.9  
• mkdir build  
• cd build  
• cmake -DWITH_CUDA=ON -DCUDA_ARCH_BIN="3.2" -DCUDA_ARCH_PTX="" -DBUILD_TESTS=OFF -DBUILD_PERF_TESTS=OFF ..  
• sudo make -j4 install  
  

④.配置环境变量

echo "# Use OpenCV and other custom-built libraries." >> ~/.bashrcecho "export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/lib/" >> ~/.bashrc
source ~/.bashrc
⑤.测试

第一个例子：边缘提取

# Test a simple OpenCV program. Creates a graphical window, hence you should plug a HDMI monitor in or use a remote viewer such as X Tunneling or VNC or TeamViewer on your desktop.cd ~/opencv-2.4.9/samples/cppg++ edge.cpp -lopencv_core -lopencv_imgproc -lopencv_highgui -o edge( Or for OpenCV 3.0: g++ edge.cpp -lopencv_core -lopencv_imgproc -lopencv_highgui -lopencv_imgcodecs -o edge)./edge 第二个例子：摄像头图像的拉普拉斯变换

# If you have a USB webcam plugged in to your board, then test one of the live camera programs and linking to every OpenCV module.g++ laplace.cpp -lopencv_core -lopencv_imgproc -lopencv_highgui -lopencv_calib3d -lopencv_contrib -lopencv_features2d -lopencv_flann -lopencv_gpu -lopencv_legacy -lopencv_ml -lopencv_objdetect -lopencv_photo -lopencv_stitching -lopencv_superres -lopencv_video -lopencv_videostab -o laplace 第三个例子：CPU和GPU进行图像的霍夫变换

# Test a GPU accelerated OpenCV sample.cd ../gpug++ houghlines.cpp -lopencv_core -lopencv_imgproc -lopencv_highgui -lopencv_calib3d -lopencv_contrib -lopencv_features2d -lopencv_flann -lopencv_gpu -lopencv_legacy -lopencv_ml -lopencv_objdetect -lopencv_photo -lopencv_stitching -lopencv_superres -lopencv_video -lopencv_videostab -o houghlines./houghlines ../cpp/logo_in_clutter.png 三、QT的安装与配置

1.安装，使用apt-get，编译opensource的方法没有通过//不过！！！貌似只安装qtcreater就行了，商店里有

sudo apt- get install qt4-dev- tools #开发包 　　sudo apt- get install qtcreator #IDE　　sudo apt- get install qt4- doc #开发帮助文档 　　sudo apt- get install qt4- qtconfig #配置工具 　　sudo apt- get install qt4-demos #DEMO源码  QT 2 .配置(这一步的截图见csdn:http://blog.csdn.net/u012348655/article/details/50564223)

①.打开qt->Tools->Options->Build&Run

②.配置Qt Versions以及Compilers

在termial中输入which qmake以及which gcc把相应的地址添加进去

3.测试

新建一个控制台工程

在main.c中加入

[cpp] view plain copy

                               
登录/注册后可看大图

                               
登录/注册后可看大图

• #include   
• #include   
• #include   
• #include   
• #include   
• #include   
•   
• using namespace cv;  
• using namespace std;  
•   
• int main(int argc, char *argv[])  
• {  
• //    QCoreApplication a(argc, argv);  
•   
•     namedWindow( "src", WINDOW_AUTOSIZE );  
•     printf("hello lena\n");  
•     Mat src1 = imread( "lena.jpg" );  
•     while(1)  
•     {  
•         imshow( "src", src1 );  
•         char c =  waitKey(30);  
•         if( 27==c )  
•             return 0;  
•     }  
•   
• //    return a.exec();  
• }  
  

在.pro文件中加入

[cpp] view plain copy

                               
登录/注册后可看大图

                               
登录/注册后可看大图

• SOURCES += main.cpp  
• INCLUDEPATH += /usr/local/include \  
•                 /usr/local/include/opencv \  
•                 /usr/local/include/opencv2  
•   
• LIBS += /usr/local/lib/libopencv_highgui.so \  
•         /usr/local/lib/libopencv_core.so    \  
•         /usr/local/lib/libopencv_imgproc.so  
  

在生成的调试文件夹中放入lena.jpg
然后运行