1.配置环境
windows10 + VisualStudio2017community + YOLOV3
cuda10.0.130_411.31
cunn-10.0-v7.6.5.32
opencv3.2
常规
windows sdk版本 10.0.19041.0（10.0.17763.0）
平台工具集 Visual Studio 2017(V141)
C++ ---常规
(CudaToolkitIncludeDir);
(cudnn)\include
VC++ 目录
----可执行文件目录
D:\opencv\opencv\build\include;D:\opencv\opencv\build\include\opencv;D:\opencv\opencv\build\include\opencv2;(IncludePath)
----库目录
D:\opencv\opencv\build\x64\vc14\lib;$(LibraryPath)
链接器---常规---附加库目录
D:\opencv\opencv\build\x64\vc14\lib;(PlatformName);
(cudnn)\lib\x64;....\3rdparty\pthreads\lib;opencv_world320.lib;%(AdditionalLibraryDirectories)
链接器---输入---附加依赖项
pthreadVC2.lib
cublas.lib
opencv_world320.lib
opencv_world320d.lib
curand.lib
cudart.lib
2.使用VS2017进行编译
3.训练
1.修改coco.data, coco.names
2.需要修改yolov3.cfg 一个是将test注释掉，train注释取消，将classes类别数设置正确，卷积核filter个数设置为3*（类别数+1+4）
darknet.exe detector train data/coco.data cfg/yolov3.cfg darknet53.conv.74
4.预测
darknet.exe detector test data/coco.data cfg/yolov3.cfg backup/yolov3_final.weights sample.jpg
5.测试
darknet.exe detector vaild data/coco.data cfg/yolov3.cfg backup/yolov3_final.weights
6.视频demo
darknet.exe detector demo data/coco.data cfg/yolov3.cfg backup/yolov3_final.weights "sample.mp4"
参考链接：
Windows10+YOLOV3+VisualStudio2017最新版本超详细过程
CUDA10.1配置VS2017
cudnn下载链接
cuda下载链接
VS2017永久配置openCV
