docker--cuda10.1-ubuntu18.04-opencv4.0-caffe

最新推荐文章于 2024-06-23 19:03:37 发布

紫晨君

最新推荐文章于 2024-06-23 19:03:37 发布

阅读量2.9k

点赞数 3

分类专栏： AI 机器学习

本文链接：https://blog.csdn.net/weixin_38665284/article/details/88812950

版权

本文详细记录了在Ubuntu18.04的Docker环境中，安装CUDA 10.1、OpenCV 4.0和Caffe的步骤，包括遇到的问题如OpenBLAS找不到库、Python3编译错误、CUDA cublas_device库缺失、CMake版本过低以及找不到GFlags、GLog和LMDB库等问题，以及相应的解决方案，如升级CMake、手动配置Caffe的Makefile和CMakeCache.txt等。

摘要由CSDN通过智能技术生成

文章目录

1 cuda镜像

# 查找docker源中的cuda镜像
docker search nvida/cuda # 此镜像默认是cuda10.0，os是ubuntu16.04
# 查询docker hub网站链接到https://gitlab.com/nvidia/cuda/blob/ubuntu18.04/10.1/devel/cudnn7/Dockerfile到dockerfile内容，发现可以定制docker进行内容，因此下载如下docker镜像
docker pull nvidia/cuda:10.1-cudnn7-devel-ubuntu18.04

2 ubuntu18.04基础环境

apt update
apt install vim -y

# ubuntu换国内源
vim /etc/apt/sources.list
# 文件头新增清华源：
deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic main restricted universe multiverse
deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic main restricted universe multiverse
deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-updates main restricted universe multiverse
deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-updates main restricted universe multiverse
deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-backports main restricted universe multiverse
deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-backports main restricted universe multiverse
deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-security main restricted universe multiverse
deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-security main restricted universe multiverse
deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-proposed main restricted universe multiverse
deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-proposed main restricted universe multiverse

apt update
apt upgrade
apt install wget git cmake -y

3 安装python3.6

apt install python3 -y
apt install python3-pip -y
pip3 install --upgrade pip
pip config set global.index-url https://pypi.tuna.tsinghua.edu.cn/simple

4 opencv4.0

# 安装依赖包
apt install build-essential libgtk2.0-dev libavcodec-dev libavformat-dev libjpeg-dev libswscale-dev libtiff5-dev -y

pip install numpy

# 下载opencv 4.0
https://opencv.org/releases.html

# 源码编译安装
cd opencv-4.0.1
mkdir build
cd build
cmake ..
make all -j8
make install

问题：

– Could not find OpenBLAS lib. Turning OpenBLAS_FOUND

apt install libopenblas-dev libopenblas-base liblapacke-dev -y

# 软连接正确的库路径
ln -s /usr/include/lapacke.h /usr/include/x86_64-linux-gnu

无法编译python3

解决方式如下：

cmake -DBUILD_LIBPROTOBUF_FROM_SOURCES=ON\
	-D CMAKE_BUILD_TYPE=RELEASE \
  -D CMAKE_INSTALL_PREFIX=/usr/local \
  -D INSTALL_C_EXAMPLES=OFF \
  -D INSTALL_PYTHON_EXAMPLES=ON \
  -D BUILD_EXAMPLES=ON \
  -D PYTHON3_EXECUTABLE=