我使用的是OpenSUSE 13.2-X86-64和Geforce 940M GPU。我想在Qt5 IDE上使用它。 为此，我已经从cudazone安装了我的GPU驱动程序和Cuda7.5工具包rpm。一切都差不多好了。 它检测我所做的一切并执行了几个cuda样本。
下一步，我已经安装了当前的FFmpeg版本与nvenc和其他库，如Xvid和许多 其他有用的东西，在这里列出很多。之后我下载了当前的 OpenCV-3.0.0源代码并运行cmake-gui，其中我添加了cuda，ffmpeg，Qt5等并制作了它。
在make会话的不同点（make -j4）我从我的控制台得到这种错误（多重定义的列表） 错误更长）。这是第一个。
.
.
.
.
.
.
nvlink error : Multiple definition of '_ZN2cv5cudev16color_cvt_detail15c_HlsSectorDataE' in '/home/peter/Programme/opencv/build/modules/cudev/test/CMakeFiles/opencv_test_cudev.dir//./opencv_test_cudev_generated_test_arithm_func.cu.o', first defined in '/home/peter/Programme/opencv/build/modules/cudev/test/CMakeFiles/opencv_test_cudev.dir//./opencv_test_cudev_generated_test_lut.cu.o'
nvlink error : Multiple definition of '_ZN2cv5cudev16color_cvt_detail16c_sRGBGammaTab_bE' in '/home/peter/Programme/opencv/build/modules/cudev/test/CMakeFiles/opencv_test_cudev.dir//./opencv_test_cudev_generated_test_arithm_func.cu.o', first defined in '/home/peter/Programme/opencv/build/modules/cudev/test/CMakeFiles/opencv_test_cudev.dir//./opencv_test_cudev_generated_test_lut.cu.o'
nvlink error : Multiple definition of '_ZN2cv5cudev16color_cvt_detail14c_sRGBGammaTabE' in '/home/peter/Programme/opencv/build/modules/cudev/test/CMakeFiles/opencv_test_cudev.dir//./opencv_test_cudev_generated_test_arithm_func.cu.o', first defined in '/home/peter/Programme/opencv/build/modules/cudev/test/CMakeFiles/opencv_test_cudev.dir//./opencv_test_cudev_generated_test_lut.cu.o'
nvlink error : Multiple definition of '_ZN2cv5cudev16color_cvt_detail17c_sRGBInvGammaTabE' in '/home/peter/Programme/opencv/build/modules/cudev/test/CMakeFiles/opencv_test_cudev.dir//./opencv_test_cudev_generated_test_arithm_func.cu.o', first defined in '/home/peter/Programme/opencv/build/modules/cudev/test/CMakeFiles/opencv_test_cudev.dir//./opencv_test_cudev_generated_test_lut.cu.o'
nvlink error : Multiple definition of '_ZN2cv5cudev16color_cvt_detail12c_LabCbrtTabE' in '/home/peter/Programme/opencv/build/modules/cudev/test/CMakeFiles/opencv_test_cudev.dir//./opencv_test_cudev_generated_test_arithm_func.cu.o', first defined in '/home/peter/Programme/opencv/build/modules/cudev/test/CMakeFiles/opencv_test_cudev.dir//./opencv_test_cudev_generated_test_lut.cu.o'
modules/cudev/test/CMakeFiles/opencv_test_cudev.dir/build.make:5302: recipe for target 'modules/cudev/test/CMakeFiles/opencv_test_cudev.dir/./opencv_test_cudev_intermediate_link.o' failed
make[2]: *** [modules/cudev/test/CMakeFiles/opencv_test_cudev.dir/./opencv_test_cudev_intermediate_link.o] Error 255
CMakeFiles/Makefile2:1182: recipe for target 'modules/cudev/test/CMakeFiles/opencv_test_cudev.dir/all' failed
make[1]: *** [modules/cudev/test/CMakeFiles/opencv_test_cudev.dir/all] Error 2
Makefile:137: recipe for target 'all' failed
make: *** [all] Error 2
我不知道如何解决这个问题。
THX！
编辑：添加了cmake配置
~/Programme/opencv/build> cmake /home/peter/Programme/opencv-3.0.0
CMake Error: The source "/home/peter/Programme/opencv-3.0.0/CMakeLists.txt" does not match the source "/home/peter/Programme/opencv/CMakeLists.txt" used to generate cache. Re-run cmake with a different source directory.
peter@linux-3mgc:~/Programme/opencv/build> cmake /home/peter/Programme/opencv
-- Detected version of GNU GCC: 48 (408)
-- Found ZLIB: /usr/lib64/libz.so (found suitable version "1.2.8", minimum required is "1.2.3")
-- Found ZLIB: /usr/lib64/libz.so (found version "1.2.8")
-- checking for module 'gstreamer-video-0.10'
-- package 'gstreamer-video-0.10' not found
-- checking for module 'gstreamer-app-0.10'
-- package 'gstreamer-app-0.10' not found
-- checking for module 'gstreamer-riff-0.10'
-- package 'gstreamer-riff-0.10' not found
-- checking for module 'gstreamer-pbutils-0.10'
-- package 'gstreamer-pbutils-0.10' not found
-- Looking for linux/videodev.h
-- Looking for linux/videodev.h - not found
-- Looking for linux/videodev2.h
-- Looking for linux/videodev2.h - found
-- Looking for sys/videoio.h
-- Looking for sys/videoio.h - not found
-- checking for module 'libavresample'
-- package 'libavresample' not found
-- Looking for libavformat/avformat.h
-- Looking for libavformat/avformat.h - found
-- Looking for ffmpeg/avformat.h
-- Looking for ffmpeg/avformat.h - not found
-- found IPP (ICV version): 8.2.1 [8.2.1]
-- at: /home/peter/Programme/opencv/3rdparty/ippicv/unpack/ippicv_lnx
-- CUDA detected: 7.5
-- CUDA NVCC target flags: -gencode;arch=compute_50,code=sm_50
-- To enable PlantUML support, set PLANTUML_JAR environment variable or pass -DPLANTUML_JAR=<filepath> option to cmake
-- Found PythonInterp: /usr/bin/python2.7 (found suitable version "2.7.8", minimum required is "2.7")
-- Found PythonLibs: /usr/lib64/libpython2.7.so (found suitable exact version "2.7.8")
-- Found PythonInterp: /usr/bin/python3.4 (found suitable version "3.4.1", minimum required is "3.4")
-- Found PythonLibs: /usr/lib64/libpython3.4m.so (found suitable exact version "3.4.1")
Traceback (most recent call last):
File "<string>", line 1, in <module>
ImportError: No module named 'numpy'
-- Found apache ant 1.8.0: /usr/bin/ant
-- Could NOT find JNI (missing: JAVA_INCLUDE_PATH JAVA_INCLUDE_PATH2 JAVA_AWT_INCLUDE_PATH)
-- Could NOT find Matlab (missing: MATLAB_MEX_SCRIPT MATLAB_INCLUDE_DIRS MATLAB_ROOT_DIR MATLAB_LIBRARIES MATLAB_LIBRARY_DIRS MATLAB_MEXEXT MATLAB_ARCH MATLAB_BIN)
-- VTK support is disabled. Incompatible combination: OpenCV + Qt5 and VTK ver.6.1.0 + Qt4
--
-- General configuration for OpenCV 3.0.0-dev =====================================
-- Version control: 3.0.0-528-g3a3f403-dirty
--
-- Platform:
-- Host: Linux 3.16.7-24-desktop x86_64
-- CMake: 3.0.2
-- CMake generator: Unix Makefiles
-- CMake build tool: /usr/bin/gmake
-- Configuration: Release
--
-- C/C++:
-- Built as dynamic libs?: YES
-- C++ Compiler: /usr/bin/c++ (ver 4.8.3)
-- C++ flags (Release): -fsigned-char -W -Wall -Werror=return-type -Werror=non-virtual-dtor -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wundef -Winit-self -Wpointer-arith -Wshadow -Wsign-promo -Wno-narrowing -Wno-delete-non-virtual-dtor -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -msse -msse2 -mno-avx -msse3 -mno-ssse3 -mno-sse4.1 -mno-sse4.2 -ffunction-sections -fvisibility=hidden -fvisibility-inlines-hidden -fopenmp -O3 -DNDEBUG -DNDEBUG
-- C++ flags (Debug): -fsigned-char -W -Wall -Werror=return-type -Werror=non-virtual-dtor -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wundef -Winit-self -Wpointer-arith -Wshadow -Wsign-promo -Wno-narrowing -Wno-delete-non-virtual-dtor -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -msse -msse2 -mno-avx -msse3 -mno-ssse3 -mno-sse4.1 -mno-sse4.2 -ffunction-sections -fvisibility=hidden -fvisibility-inlines-hidden -fopenmp -g -O0 -DDEBUG -D_DEBUG
-- C Compiler: /usr/bin/cc
-- C flags (Release): -fsigned-char -W -Wall -Werror=return-type -Werror=non-virtual-dtor -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wmissing-prototypes -Wstrict-prototypes -Wundef -Winit-self -Wpointer-arith -Wshadow -Wno-narrowing -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -msse -msse2 -mno-avx -msse3 -mno-ssse3 -mno-sse4.1 -mno-sse4.2 -ffunction-sections -fvisibility=hidden -fopenmp -O3 -DNDEBUG -DNDEBUG
-- C flags (Debug): -fsigned-char -W -Wall -Werror=return-type -Werror=non-virtual-dtor -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wmissing-prototypes -Wstrict-prototypes -Wundef -Winit-self -Wpointer-arith -Wshadow -Wno-narrowing -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -msse -msse2 -mno-avx -msse3 -mno-ssse3 -mno-sse4.1 -mno-sse4.2 -ffunction-sections -fvisibility=hidden -fopenmp -g -O0 -DDEBUG -D_DEBUG
-- Linker flags (Release):
-- Linker flags (Debug):
-- Precompiled headers: YES
-- Extra dependencies: /usr/lib64/libcuda.so /usr/lib64/libnvcuvid.so Qt5::Core Qt5::Gui Qt5::Widgets Qt5::Test Qt5::Concurrent Qt5::OpenGL /usr/lib64/libwebp.so /usr/lib64/libpng.so /usr/lib64/libz.so /usr/lib64/libtiff.so /usr/lib64/libjasper.so /usr/lib64/libjpeg.so gstbase-0.10 gstreamer-0.10 gobject-2.0 gmodule-2.0 gthread-2.0 xml2 ucil glib-2.0 unicap dc1394 xine v4l1 v4l2 avcodec avformat avutil swscale gphoto2 gphoto2_port exif /usr/lib64/libbz2.so dl m pthread rt /usr/lib64/libGLU.so /usr/lib64/libGL.so /usr/lib64/libSM.so /usr/lib64/libICE.so /usr/lib64/libX11.so /usr/lib64/libXext.so cudart nppc nppi npps cublas cufft
-- 3rdparty dependencies: IlmImf ippicv
--
-- OpenCV modules:
-- To be built: cudev hal core cudaarithm flann imgproc ml video cudabgsegm cudafilters cudaimgproc cudawarping imgcodecs photo shape videoio cudacodec highgui objdetect ts features2d calib3d cudafeatures2d cudalegacy cudaobjdetect cudaoptflow cudastereo stitching superres videostab python2
-- Disabled: world
-- Disabled by dependency: -
-- Unavailable: java python3 viz
--
-- GUI:
-- QT 5.x: YES (ver 5.4.2)
-- QT OpenGL support: YES (Qt5::OpenGL 5.4.2)
-- OpenGL support: YES (/usr/lib64/libGLU.so /usr/lib64/libGL.so /usr/lib64/libSM.so /usr/lib64/libICE.so /usr/lib64/libX11.so /usr/lib64/libXext.so)
-- VTK support: NO
--
-- Media I/O:
-- ZLib: /usr/lib64/libz.so (ver 1.2.8)
-- JPEG: /usr/lib64/libjpeg.so (ver )
-- WEBP: /usr/lib64/libwebp.so (ver encoder: 0x0202)
-- PNG: /usr/lib64/libpng.so (ver 1.2.51)
-- TIFF: /usr/lib64/libtiff.so (ver 42 - 4.0.4)
-- JPEG 2000: /usr/lib64/libjasper.so (ver 1.900.1)
-- OpenEXR: build (ver 1.7.1)
-- GDAL: NO
--
-- Video I/O:
-- DC1394 1.x: NO
-- DC1394 2.x: YES (ver 2.2.2)
-- FFMPEG: YES
-- codec: YES (ver 57.3.100)
-- format: YES (ver 57.2.100)
-- util: YES (ver 55.2.100)
-- swscale: YES (ver 4.0.100)
-- resample: NO
-- gentoo-style: YES
-- GStreamer: NO
-- OpenNI: NO
-- OpenNI PrimeSensor Modules: NO
-- OpenNI2: NO
-- PvAPI: NO
-- GigEVisionSDK: NO
-- UniCap: YES (ver 0.9.12)
-- UniCap ucil: YES (ver 0.9.10)
-- V4L/V4L2: Using libv4l1 (ver 1.2.1) / libv4l2 (ver 1.2.1)
-- XIMEA: NO
-- Xine: YES (ver 1.2.6)
-- gPhoto2: YES
--
-- Parallel framework: OpenMP
--
-- Other third-party libraries:
-- Use IPP: 8.2.1 [8.2.1]
-- at: /home/peter/Programme/opencv/3rdparty/ippicv/unpack/ippicv_lnx
-- Use IPP Async: NO
-- Use VA: NO
-- Use Intel VA-API/OpenCL: NO
-- Use Eigen: YES (ver 3.2.2)
-- Use Cuda: YES (ver 7.5)
-- Use OpenCL: YES
--
-- NVIDIA CUDA
-- Use CUFFT: YES
-- Use CUBLAS: YES
-- USE NVCUVID: YES
-- NVIDIA GPU arch: 50
-- NVIDIA PTX archs:
-- Use fast math: YES
--
-- OpenCL:
-- Version: dynamic
-- Include path: /home/peter/Programme/opencv/3rdparty/include/opencl/1.2
-- Use AMDFFT: NO
-- Use AMDBLAS: NO
--
-- Python 2:
-- Interpreter: /usr/bin/python2.7 (ver 2.7.8)
-- Libraries: /usr/lib64/libpython2.7.so (ver 2.7.8)
-- numpy: /usr/lib64/python2.7/site-packages/numpy/core/include (ver 1.9.0)
-- packages path: lib/python2.7/site-packages
--
-- Python 3:
-- Interpreter: /usr/bin/python3.4 (ver 3.4.1)
--
-- Python (for build): /usr/bin/python2.7
--
-- Java:
-- ant: /usr/bin/ant (ver 1.8.0)
-- JNI: NO
-- Java wrappers: NO
-- Java tests: NO
--
-- Matlab:
-- mex: NO
--
-- Documentation:
-- Doxygen: /usr/bin/doxygen (ver 1.8.8)
-- PlantUML: NO
--
-- Tests and samples:
-- Tests: YES
-- Performance tests: YES
-- C/C++ Examples: NO
--
-- Install path: /usr/local
--
-- cvconfig.h is in: /home/peter/Programme/opencv/build
--
-----------------------------------------------------------------
--
-- Configuring done
-- Generating done
-- Build files have been written to: /home/peter/Programme/opencv/build
答案 0 :(得分：1)
所以在尝试并再次尝试后，我以不同的方式解决了问题。我从opencv3退回到2.4.11，我退出了最新的ffmpeg（这解决了很多问题）。另一个问题是安装的openv-2.4.9版本总是希望使用而不是我的cuda编译库。我不是一个想要阅读cmakefiles的人。 OpenCV有很多这样的东西，所以cmake肯定会让我感到困惑。
尽管如此，我已经解决了我的问题。不是使用OpenCV3，而是使用opencv-2.4.11，Cuda-5.0，这对我来说已经足够了。为了运行这个组合，我写了一些shell脚本来反馈我自己并分析问题。此文件应以超级用户身份运行，并替换用户名，当然也可以在自定义系统中安装文件。
- 请注意，您可能需要重新启动bumbleed守护程序，使用mkinitrd并重新启动另一次。由于这里的指令，第二次强制安装完成了 - ＆gt; https://en.opensuse.org/SDB:NVIDIA_Bumblebee 在结束部分。如果你想使用cuda开发，建议。 在此之后，您必须重新启动才能完成它。
mkdir /home/peter/programs
cd /home/peter/programs
# 3rd part library and qt library packages installed. Not even sure that everyone is needed
zypper --non-interactive in gcc git libwebp-devel vtk vtk-devel eigen3-devel libeigen2-devel python-numpy-devel python python3-numpy python3-numpy-devel ant java-1_7_0-openjdk-devel java-1_7_0-openjdk-src python-Sphinx openexr openexr-devel libpng12-devel openjpeg openjpeg-devel tiff jasper libjasper-devel pbzip2 python3-cairo python3-cairo-devel ilmbase-devel cmake-gui gtk2-devel libdc1394 libdc1394-devel libv4l-devel libxine-devel gstreamer-0_10-devel gstreamer-devel openmpi openmpi-devel openmpi-libs VirtualGL-devel Mesa-libGLESv3-devel ftgl-devel glew glproto-devel gltt gltt-devel gtkglext-devel libGLw1 libftgl2 libglfw-devel libgltf-devel python-opengl gluegen2 gluegen2-devel libjnidispatch gphoto libgphoto2-devel libunicap2 libunicap-devel libucil-devel libucil2 zypper libbz2-devel doxygen jna zypper libQt5Bluetooth5 libQt5Bootstrap-devel-static libQt5Concurrent-devel libQt5Concurrent5 libQt5Core-devel libQt5Core-private-headers-devel libQt5Designer5 libQt5DesignerComponents5 libQt5Gui-devel libQt5Gui-private-headers-devel libQt5Help5 libQt5Network-devel libQt5Network-private-headers-devel libQt5Network5 libQt5OpenGL-devel libQt5OpenGL-private-headers-devel libQt5OpenGL5 libQt5OpenGLExtensions-devel-static libQt5PlatformHeaders-devel libQt5PlatformSupport-devel-static libQt5PlatformSupport-private-headers-devel libQt5Widgets-devel libqt5-creator libqt5-qtdoc libqt5-qttools libqt5-qttools-devel python-qt5 python-qt5-devel python-qt5-utils freetype freetype-devel freetype2-devel dpic libvpx1 libvpx-devel nasm yasm libopus-devel autoconf automake libass-devel libfreetype6 libSDL-devel libSDL2-devel libtheora-devel libtool libva1 libva-devel libtool libvdpau-devel libvorbis-devel libxcb-devel libxcb-shm0 libxcb-xfixes0 pkg-config latex2html zlib zlib-devel
# downloading and preparing libs that should be installed form the scratch
wget http://developer.download.nvidia.com/compute/cuda/repos/opensuse132/x86_64/cuda-repo-opensuse132-7.5-18.x86_64.rpm
mkdir SDK
cd SDK
wget http://developer.download.nvidia.com/compute/nvenc/v5.0/nvenc_5.0.1_sdk.zip -O nvenc-sdk.zip
unzip nvenc-sdk.zip
cd nvenc_5.0.1_sdk
cd /home/peter/programs
wget http://download.videolan.org/pub/x264/snapshots/last_x264.tar.bz2
tar xjvf last_x264.tar.bz2
wget -O fdk-aac.zip https://github.com/mstorsjo/fdk-aac/zipball/master
unzip fdk-aac.zip
wget http://downloads.sourceforge.net/project/lame/lame/3.99/lame-3.99.5.tar.gz
tar xzvf lame-3.99.5.tar.gz
wget http://downloads.xvid.org/downloads/xvidcore-1.3.4.tar.gz
tar xzvf xvidcore-1.3.4.tar.gz
wget http://ffmpeg.org/releases/ffmpeg-2.6.1.tar.bz2
tar xjvf ffmpeg-2.6.1.tar.bz2
wget http://sourceforge.net/projects/opencvlibrary/files/opencv-unix/2.4.11/opencv-2.4.11.zip/download --output-document=opencv.zip
unzip opencv.zip
cd opencv-2.4.11
mkdir build
cd /home/peter/programs
rpm -i cuda-repo-opensuse132-7.5-18.x86_64.rpm
zypper --gpg-auto-import-keys refresh
zypper addrepo http://download.opensuse.org/repositories/X11:Bumblebee/openSUSE_13.2/X11:Bumblebee.repo
zypper --gpg-auto-import-keys refresh
zypper --non-interactive --gpg-auto-import-keys in bumblebee
usermod -G video,bumblebee peter
systemctl enable bumblebeed
mkinitrd
zypper --gpg-auto-import-keys --non-interactive in cuda
systemctl enable dkms
zypper --non-interactive in nvidia-bumblebee-32bit
zypper --gpg-auto-import-keys --non-interactive in cuda
zypper --non-interactive in cuda-gdb-src-7-5
zypper --gpg-auto-import-keys --non-interactive in -f nvidia-bumblebee
cd /etc/ld.so.conf.d
echo -e "/usr/local/cuda-7.5/lib64" > cudald.conf
ldconfig
cd /home/peter
echo -e "export LD_LIBRARY_PATH=/usr/local/cuda-7.5/lib64:$LD_LIBRARY_PATH\nexport PATH=/usr/local/cuda-7.5/bin:$PATH" >> .bashrc
chmod 777 -R programs
reboot
我不确定它是否是问题的根源，但重新启动后，您可能会打开一个新的shell并运行此文件。再次不要忘记自定义shell脚本。在开始之前，您应该扩展您的sudo会话或再次以超级用户身份运行它。
删除rpms可能附带的任何其他opencv-libs。然后，您可以启动此代码段进行安装。
- 请注意最后需要optirun命令才能在所有步骤中使用正确的体系结构。配置有时会通过make或make install重新启动。要保持配置，请使用optimus运行。
cd /home/peter/programs/SDK/nvenc_5.0.1_sdk
sudo cp Samples/common/inc/* /usr/include/
cd /home/peter/programs/x264-snapshot*
./configure --enable-static --enable-shared --enable-pic
make
sudo make install
make distclean
cd /home/peter/programs/mstorsjo-fdk-aac*
autoreconf -fiv
./configure
make
sudo make install
make distclean
cd /home/peter/programs/lame-3.99.5
./configure --enable-nasm
make
sudo make install
make distclean
cd /home/peter/programs/xvidcore/build/generic
./configure
make
sudo make install
make distclean
cd /home/peter/programs/ffmpeg-2.6.1
./configure \
--enable-shared \
--enable-gpl \
--enable-libass \
--enable-libfdk-aac \
--enable-libfreetype \
--enable-libmp3lame \
--enable-libopus \
--enable-libtheora \
--enable-libvorbis \
--enable-libvpx \
--enable-libx264 \
--enable-libxvid \
--enable-nonfree \
--enable-nvenc
make -j8
sudo make install
make distclean
hash -r
echo "Complete!"
cd /home/peter/programs/opencv-2.4.11/build
optirun cmake -D BUILD_EXAMPLES:BOOL=TRUE -D CUDA_ARCH_BIN:STRING=5.0 -D CUDA_ARCH_PTX:STRING=5.0 -D CUDA_BUILD_CUBIN:BOOL=TRUE -D CUDA_FAST_MATH:BOOL=TRUE -D CUDA_GENERATION:STRING=Auto -D INSTALL_C_EXAMPLES:BOOL=TRUE -D INSTALL_PYTHON_EXAMPLES:BOOL=TRUE -D WITH_CUDA:BOOL=ON -D INSTALL_TESTS:BOOL=TRUE -D WITH_CUBLAS:BOOL=TRUE -D WITH_NVCUVID:BOOL=TRUE -D WITH_OPENCLAMDBLAS:BOOL=FALSE -D WITH_OPENCLAMDFFT:BOOL=FALSE -D WITH_OPENGL:BOOL=TRUE -D WITH_UNICAP:BOOL=TRUE -D WITH_XINE:BOOL=TRUE -D WITH_QT:BOOL=TRUE -D CUDA_nvcuvid_LIBRARY:FILEPATH=/usr/lib64/libnvcuvid.so.1 /home/peter/programs/opencv-2.4.11/
optirun make -j8
sudo optirun make install
使用optirun启动qtcreator，你应该能像其他所有opencv程序一样编译和运行opencv_gpu文件。
Thx peter