从上一篇的Tensorflow win10 c++ 运行 python训练出的模型，按照Tensorflow官网给出的cmake构建和编译方案，我们实际编译了tensorflow的c++库，能够运行官方的example。那么新建一个单独的工程需要进行一定的配置。

源代码使用官方的源码main.cc

这份代码包含了读取模型，读取图片数据，进行模型预测等，足够满足简单的tensorflow功能需求。
注意的是，需要在代码头部加上

#define COMPILER_MSVC
#define NOMINMAX

原因如这篇博客所述

If you omit the COMPILER_MSVC definition, you will run into an error saying “You must define TF_LIB_GTL_ALIGNED_CHAR_ARRAY for your compiler.” If you omit the NOMINMAX definition, you will run into a number of errors saying “’(‘: illegal token on right side of ‘::’”. (The reason for this is that <Windows.h> gets included somewhere, and Windows has macros that redefine min and max. These macros are disabled with NOMINMAX.)

工程属性设置

接下来配置文件，首先本文同步的tensorflow源代码位置为D:\Projects\tensorflow。

附加包含路径

设置对应如下的包含路径，可以通过直接编辑官方例子tf_label_image_example.vcxproj到自己的工程文件

D:\Projects\tensorflow
D:\Projects\tensorflow\tensorflow\contrib\cmake\build
D:\Projects\tensorflow\tensorflow\contrib\cmake\build\external\zlib_archive
D:\Projects\tensorflow\tensorflow\contrib\cmake\build\external\gif_archive\giflib-5.1.4
D:\Projects\tensorflow\tensorflow\contrib\cmake\build\external\png_archive
D:\Projects\tensorflow\tensorflow\contrib\cmake\build\external\jpeg_archive
D:\Projects\tensorflow\tensorflow\contrib\cmake\build\external\eigen_archive
D:\Projects\tensorflow\third_party\eigen3
D:\Projects\tensorflow\tensorflow\contrib\cmake\build\gemmlowp\src\gemmlowp
D:\Projects\tensorflow\tensorflow\contrib\cmake\build\jsoncpp\src\jsoncpp
D:\Projects\tensorflow\tensorflow\contrib\cmake\build\external\farmhash_archive
D:\Projects\tensorflow\tensorflow\contrib\cmake\build\external\farmhash_archive\util
D:\Projects\tensorflow\tensorflow\contrib\cmake\build\external\highwayhash
D:\Projects\tensorflow\tensorflow\contrib\cmake\build\protobuf\src\protobuf\src
D:\Projects\tensorflow\tensorflow\contrib\cmake\build\grpc\src\grpc\include

链接设置

按照tf_label_image_example.vcxproj添加依赖项目，按照博客添加额外依赖路径

kernel32.lib
user32.lib
gdi32.lib
winspool.lib
shell32.lib
ole32.lib
oleaut32.lib
uuid.lib
comdlg32.lib
advapi32.lib
Release\tf_protos_cc.lib
zlib\install\lib\zlibstatic.lib
gif\install\lib\giflib.lib
png\install\lib\libpng12_static.lib
jpeg\install\lib\libjpeg.lib
jsoncpp\src\jsoncpp\src\lib_json\$(Configuration)\jsoncpp.lib
farmhash\install\lib\farmhash.lib
fft2d\\src\lib\fft2d.lib
highwayhash\install\lib\highwayhash.lib
protobuf\src\protobuf\$(Configuration)\libprotobuf.lib
grpc\src\grpc\$(Configuration)\grpc++_unsecure.lib
grpc\src\grpc\$(Configuration)\grpc_unsecure.lib
grpc\src\grpc\$(Configuration)\gpr.lib
wsock32.lib
ws2_32.lib
shlwapi.lib

D:\Projects\tensorflow\tensorflow\contrib\cmake\build\protobuf\src\protobuf\Release
D:\Projects\tensorflow\tensorflow\contrib\cmake\build\tf_cc.dir\Release
D:\Projects\tensorflow\tensorflow\contrib\cmake\build\tf_cc_ops.dir\Release
D:\Projects\tensorflow\tensorflow\contrib\cmake\build\tf_cc_framework.dir\Release
D:\Projects\tensorflow\tensorflow\contrib\cmake\build\tf_core_cpu.dir\Release
D:\Projects\tensorflow\tensorflow\contrib\cmake\build\tf_core_direct_session.dir\Release
D:\Projects\tensorflow\tensorflow\contrib\cmake\build\tf_core_framework.dir\Release
D:\Projects\tensorflow\tensorflow\contrib\cmake\build\tf_core_kernels.dir\Release
D:\Projects\tensorflow\tensorflow\contrib\cmake\build\tf_core_lib.dir\Release
D:\Projects\tensorflow\tensorflow\contrib\cmake\build\tf_core_ops.dir\Release
D:\Projects\tensorflow\tensorflow\contrib\cmake\build\Release
D:\Projects\tensorflow\tensorflow\contrib\cmake\build

添加预编译好的obj文件，在tf_label_image_example.vcxproj文件中找到
<Object Include="D:\Projects\tensorflow\tensorflow\contrib\cmake\build\tf_core_lib.dir\$(Configuration)\arena.obj" />
开头的一长串语句，复制到我们工程目录中vcxproj的相应位置。

至此，程序就可以编译调试了。