使用OpenCv 2.4.10 + Cuda 7.5 + W10不支持GPU

No GPU Support using OpenCv 2.4.10 + Cuda 7.5 + W10

本文关键字：W10 不支持 GPU Cuda 使用 OpenCv      更新时间：2023-10-16

我试图让GPU启动和运行使用:

  • OpenCv 2.4.10
  • Visual Studio 2013
  • Cuda工具包7.5 (Cuda样本测试和运行)
  • Windows 10 x64
  • 联想Y50配NVIDIA GEFORCE GTX 860M

我使用CMake with WITH_CUDA=ON编译了OpenCv，然后通过打开OpenCv安装了库。

设置我的项目配置为:

  • vc++目录
    • 包括目录:
      • C:Program FilesNVIDIA GPU Computing ToolkitCUDAv7.5include
      • C:OpenCv OpenCv buildinstallinclude(包含编译后的库)
      • C: OpenCv OpenCv 制造包括

  • Linker-general-Additional库
    • C: OpenCv OpenCv 制造 x64 vc12 lib
    • C: OpenCv OpenCv 制造安装 x64 vc12 lib

  • Linker-input-Additional Depencies

    • opencv_calib3d2410d.lib
    • opencv_contrib2410d.lib
    • opencv_core2410d.lib
    • opencv_features2d2410d.lib
    • opencv_flann2410d.lib
    • opencv_gpu2410d.lib
    • opencv_highgui2410d.lib
    • opencv_imgproc2410d.lib
    • opencv_legacy2410d.lib
    • opencv_ml2410d.lib
    • opencv_nonfree2410d.lib
    • opencv_objdetect2410d.lib
    • opencv_ocl2410d.lib
    • opencv_photo2410d.lib
    • opencv_stitching2410d.lib
    • opencv_superres2410d.lib
    • opencv_ts2410d.lib
    • opencv_video2410d.lib
    • opencv_videostab2410d.lib

环境变量:路径= C:Program FilesNVIDIA GPU Computing;ToolkitCUDAv7.5;C:OpenCv OpenCv buildinstallx64vc12bin;C:OpenCv OpenCv buildx64vc12bin

测试代码: 

   #include "opencv2opencv.hpp"
   #include "opencv2gpugpu.hpp"
    using namespace cv::gpu;
    using namespace std;
    int main() {
        {
        setDevice(0); 
        int gpu = getCudaEnabledDeviceCount();
        cout << gpu ;
        system("pause");
        return 0;
        }
}in

但是setDevice抛出

没有GPU支持(库在没有CUDA支持的情况下编译)EmptyDeviceInfoFuncTable:: setDevice、文件c: 2 _4_packslave-win64-vc12-shared构建 opencv dynamicuda 包括 opencv2/dynamiccuda/dynamicuda.hpp。第84行

，如果我评论setDevice(0) gpu仍然没有检测到，getCudaEnabledDeviceCount返回0。

问题是Opencv 2.4.10支持CUDA 4.2，我的显卡只兼容CUDA 6.5和7.5，我通过安装Opencv 3.0.0解决

相关文章：