1. OpenCV3

opencv3 與cuda8不相容，不啟動支援cuda
真的是歷經千辛啊

• 安裝依賴庫

sudo apt-get install cmake git pkg-config libgtk2.0-dev libavcodec-dev libavformat-dev libswscale-dev  

sudo apt-get install python-dev python-openCV python-numpy libtbb2 libtbb-dev libjpeg-dev libpng-dev libtiff-dev libjasper-dev libdc1394-
 
22-dev

• build

mkdir build
cd build
cmake -D CMAKE_BUILD_TYPE=RELEASE -D CMAKE_INSTALL_PREFIX=/usr/local ..

cmake成功後，會出現如下結果，提示配置和生成成功：
– Configuring done
– Generating done
由於CUDA 8.0不支援OpenCV的 GraphCut 演算法，可能出現以下錯誤：

~/opencv3.1.0/modules/cudalegacy/src/graphcuts.cpp:120:54: error: 'NppiGraphcutState'
 
 has not been declared
typedef NppStatus (*init_func_t)(NppiSize oSize, NppiGraphcutState** ppStat
^
/home/usrname/opencv-3.1.0/modules/cudalegacy/src/graphcuts.cpp:135:18: error: 'NppiGraphcutState' does not name a type
operator NppiGraphcutState*()
^

【解決辦法】進入opencv-3.1.0/modules/cudalegacy/src/目錄，修改graphcuts.cpp檔案，將：

#include "precomp.hpp"
#if !defined (HAVE_CUDA) || defined (CUDA_DISABLER)

改為

#include "precomp.hpp"
#if !defined (HAVE_CUDA) || defined (CUDA_DISABLER) || (CUDART_VERSION >= 8000)

然後make編譯就可以了

make -j8
sudo make install

2. Caffe

2.0 下載

2.1 Others dependencies

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 libatlas-base-dev
sudo apt-get install libhdf5-serial-dev
sudo apt-get install libgflags-dev libgoogle-glog-dev liblmdb-dev

2.2 python

sudo apt-get install python3-dev python3-pip ipython3

在 caffe-master/python，如果是python2執行

for req in $(cat requirements.txt); do sudo pip
install $req; done

如果是python3，則把pip改為pip3

for req in $(cat requirements.txt); do pip3 install $req; done

2.3 make

1. 修改
   Makefile.config, 主要為：
   
   • 去註釋： CuDNN, USE_OPENCV,OPENCV_VERSION := 3, 允許python3的include,lib
   • 新增hdf5的include,lib

下面是我用 diff Makefile.config Makefile.config.example 比較出來的結果

<  USE_CUDNN := 1
---
> # USE_CUDNN := 1
11c11
<  USE_OPENCV := 1
---
> # USE_OPENCV := 0
21c21
<  OPENCV_VERSION := 3
---
> # OPENCV_VERSION := 3
69c69
<  PYTHON_INCLUDE := $(ANACONDA_HOME)/include \
---
> # PYTHON_INCLUDE := $(ANACONDA_HOME)/include \
74,76c74,76
<  PYTHON_LIBRARIES := boost_python3 python3.5m
<  PYTHON_INCLUDE := /usr/include/python3.5m \
<                  /usr/lib/python3.5/dist-packages/numpy/core/include
---
> # PYTHON_LIBRARIES := boost_python3 python3.5m
> # PYTHON_INCLUDE := /usr/include/python3.5m \
> #                 /usr/lib/python3.5/dist-packages/numpy/core/include
90,91c90,91
< INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include /usr/include/hdf5/serial
< LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib /usr/lib/x86_64-linux-gnu/hdf5/serial
---
> INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include
> LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib
---

1. make

make all -j8
make test -j8
make runtest -j8

runtest可能會遇到錯誤
libcudart.so.6.5: cannot open shared object file: No such file or directory.
執行以下操作

sudo cp /usr/local/cuda-8.0/lib64/libcudart.so.8.0 /usr/local/lib/libcudart.so.8.0 && sudo ldconfig

sudo cp /usr/local/cuda-8.0/lib64/libcublas.so.8.0 /usr/local/lib/libcublas.so.8.0 && sudo ldconfig

sudo cp /usr/local/cuda-8.0/lib64/libcurand.so.8.0 /usr/local/lib/libcurand.so.8.0 && sudo ldconfig

sudo cp /usr/local/cuda-8.0/lib64/libcudnn.so.5 /usr/local/lib/libcudnn.so.5 && sudo ldconfig

1. make pycaffe

make pycaffe

如果遇到錯誤 /usr/bin/ld: cannot find -lboost_python3

執行

cd /usr/lib/x86_64-linux-gnu/
sudo ln -s libboost_python-py35.so libboost_python3.so

python3.5的datautil/rrule.py裡的raiseError有語法錯誤，要自己手動修改，把raiseError後面的語句用括號括起來

設定python caffe的路徑

cd caffe-master/python
pwd

sudo gedit ~/.bashrc
新增
export PYTHONPATH=/home/chuny/caffe-master/python:$PYTHONPATH
儲存後
source ~/.bashrc

【注意】
編譯一次caffe,只能支援python2或python3
安裝python的時候，用sudo pip install是安裝在根目錄下，直接用pip安裝的是在當前使用者的home裡面，其他使用者用不了

安裝caffe的時候
make all
make test
make runtest出現這樣的問題

發現sudo make runtest就沒有問題
而且python

import caffe
caffe.set_mode_gpu() 也報錯 unable to create cublas handle
改為 sudo python
import caffe
caffe.set_mode_gpu()就沒問題
所以我全部重灌,加sudo
make clean
sudo make all -j12
sudo make test -j12
sudo make runtest -j12
sudo make pycaffe -j12