分享3：ubuntu安装Caffe，及相关安装错误的解决

现在很多同学都用着python一类的框架，例如Google的Tensorflow、Facebook的Pytorch、Amazon的MXNet等，而不愿意学习caffe,其实在做图像这一块，还是caffe比较的友好，知道大家都嫌caffe编译起来麻烦，其实caffe的编译安装真的很简单，接下来我就来分享一下安装caffe的过程

环境介绍：

笔者的环境是：
ubuntu16.04
cuda8.0
caffe安装：

1. 依赖环境的安装：

sudo apt-get install libprotobuf-dev libleveldb-dev libsnappy-dev libopencv-dev libhdf5-serial-dev protobuf-compiler
sudo apt-get install --no-install-recommends libboost-all-dev
sudo apt-get install libopenblas-dev liblapack-dev libatlas-base-dev
sudo apt-get install libgflags-dev libgoogle-glog-dev liblmdb-dev
sudo apt-get install git cmake build-essential

1. 下载caffe
   在安装路径下clone,一般默认会安装在home文件加下，文件夹名称为caffe.
   笔者就是放在home路径下的
   clone地址如下：

git clone https://github.com/BVLC/caffe.git

1. 修改 Makefile.config文件

cd caffe
sudo cp Makefile.config.example Makefile.config

复制一份的原因是编译 caffe 时需要的是 Makefile.config 文件，而Makefile.config.example 只是caffe 给出的配置文件例子，不能用来编译 caffe
然后修改 Makefile.config 文件，在 caffe 目录下打开该文件：

sudo gedit Makefile.config

修改 Makefile.config 文件内容：
使用 cudnn

将#USE_CUDNN := 1
修改成：
USE_CUDNN := 1

应用 opencv 且版本是3的

将#WITH_PYTHON_LAYER := 1
修改为
WITH_PYTHON_LAYER := 1

修改 python 路径：

#Whatever else you find you need goes here. 下面的
INCLUDE_DIRS:= $(PYTHON_INCLUDE) /usr/local/include
LIBRARY_DIRS:= $(PYTHON_LIB) /usr/local/lib /usr/lib
修改为：
INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/include/hdf5/serial /usr/local/include
LIBRARY_DIRS := $(PYTHON_LIB) /usr/lib/x86_64-linux-gnu/hdf5/serial /usr/local/lib /usr/lib

修改cuda,笔者为cuda8.0，从下面的介绍信息可以看到，cuda<8.0的需要将60和61删除，大家根据自己的cuda版本来删除对应的

# CUDA architecture setting: going with all of them.
# For CUDA < 6.0, comment the *_50 through *_61 lines for compatibility.
# For CUDA < 8.0, comment the *_60 and *_61 lin Makefile.config es for compatibility.
# For CUDA >= 9.0, comment the *_20 and *_21 lines for compatibility.
CUDA_ARCH :=-gencode arch=compute_20,code=sm_20 \
-gencode arch=compute_20,code=sm_21 \-gencode arch=compute_30,code=sm_30 \
-gencode arch=compute_35,code=sm_35 \
-gencode arch=compute_50,code=sm_50 \
-gencode arch=compute_52,code=sm_52 \
-gencode arch=compute_60,code=sm_60 \
-gencode arch=compute_61,code=sm_61 \-gencode arch=compute_61,code=compute_61

因为笔者的cuda版本为8.0所以，笔者这里删除了60和61,结果如下

# CUDA architecture setting: going with all of them.
# For CUDA < 6.0, comment the *_50 through *_61 lines for compatibility.
# For CUDA < 8.0, comment the *_60 and *_61 lines for compatibility.
# For CUDA >= 9.0, comment the *_20 and *_21 lines for compatibility.
CUDA_ARCH :=-gencode arch=compute_20,code=sm_20 \
-gencode arch=compute_20,code=sm_21 \-gencode arch=compute_30,code=sm_30 \
-gencode arch=compute_35,code=sm_35 \
-gencode arch=compute_50,code=sm_50 \
-gencode arch=compute_52,code=sm_52 \
-gencode arch=compute_61,code=compute_61

1. 修改Makefile文件

sudo gedit Makefile

将：
LIBRARIES += glog gflags protobuf boost_system boost_filesystem m hdf5_hl hdf5
改为：
LIBRARIES += glog gflags protobuf boost_system boost_filesystem m hdf5_serial_hl hdf5_serial

1. 编译

make all

make test

make runtest

make pycaffe -j8

6. 测试
ctrl+alt+t 打开命令行

输入

python

再继续输入

import caffe

如果出现以下情况则表明安装成功：

1. 容易出错的地方一：
   
   这一步错误的原因是，我们在编译之前修改 Makefile.config时，没有修改正确，应该根据提示进行修改：

# CUDA architecture setting: going with all of them.
# For CUDA < 6.0, comment the *_50 through *_61 lines for compatibility.
# For CUDA < 8.0, comment the *_60 and *_61 lines for compatibility.
# For CUDA >= 9.0, comment the *_20 and *_21 lines for compatibility.
CUDA_ARCH :=-gencode arch=compute_20,code=sm_20 \
-gencode arch=compute_20,code=sm_21 \-gencode arch=compute_30,code=sm_30 \
-gencode arch=compute_35,code=sm_35 \
-gencode arch=compute_50,code=sm_50 \
-gencode arch=compute_52,code=sm_52 \
-gencode arch=compute_60,code=sm_60 \
-gencode arch=compute_61,code=sm_61 \-gencode arch=compute_61,code=compute_61

例如：你的CUDA版本小于6.0删除如下两行：

-gencode arch=compute_50,code=sm_50 \
-gencode arch=compute_61,code=sm_61 \

如果你的CUDA版本小于8.0则删除如下两行：

-gencode arch=compute_60,code=sm_60 \
-gencode arch=compute_61,code=sm_61 \

如果你的CUDA版本小于9.0则删除如下两行：

-gencode arch=compute_20,code=sm_20 \
-gencode arch=compute_20,code=sm_21 \

1. 容易出错的地方二：
   ./include/caffe/util/hdf5.hpp:7:18: fatal error: hdf5.h: No such file or directory

In file included from src/caffe/solvers/sgd_solver.cpp:5:0:
./include/caffe/util/hdf5.hpp:7:18: fatal error: hdf5.h: No such file or directory
compilation terminated.
Makefile:591: recipe for target '.build_release/src/caffe/solvers/sgd_solver.o' failed
make: *** [.build_release/src/caffe/solvers/sgd_solver.o] Error

发现是依赖出问题了，我已经安装了hdf5了
修改Makefile.config

sudo gedit Makefile.config

修改如下：

# Whatever else you find you need goes here.
INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/include/hdf5/serial /usr/local/include
LIBRARY_DIRS := $(PYTHON_LIB) /usr/lib/x86_64-linux-gnu/hdf5/serial /usr/local/lib /usr/lib

总结：

到这里安装就结束了，看得出来，安装caffe的过程不是太复杂，知道大家跟着我的步骤来，基本上都能安装成功！！
有什么问题欢迎交流：1017190168