Ubuntu+caffe+conda+opencv+anaconda

安装conda,要保证你的GPU事大于3.1的计算能力。否则别继续看了

1,一个干净的ubuntu16.0.4系统。
2,,安装请按照(https://zhuanlan.zhihu.com/p/25709284)一步一步来。如果出错再来这里找答案就好。

解决问题

1.sudo apt-get install –no-install-recommends libboost-all-dev 有问题
解决:(http://blog.csdn.net/donatellobzero/article/details/51304162)
sudo apt-get install libboost1.55-all-dev

2.显卡安装
解决:(www.cnblogs.com/xinyf/p/5960480.html)

3.sudo ldconfig
/sbin/ldconfig.real: /usr/lib/nvidia-375/libEGL.so.1 不是符号连接
/sbin/ldconfig.real: /usr/lib32/nvidia-375/libEGL.so.1 不是符号连接
解决:http://blog.csdn.net/u014696921/article/details/70101461

4.cd /home/gl/NVIDIA_CUDA-8.0_Samples找不到目录
解决: 安装CUDA自带的Samples 
cd /usr/local/cuda/samples 
sudo make all -j4
编译完之后
cd ./bin/x86_64/linux/release
运行例程
./deviceQuery 

5.无法定位软件包
解决:–assume-yes(删除)


6.CMakeFiles/Makefile2:9117: recipe for target 'modules/cudalegacy/CMakeFiles/opencv_cudalegacy.dir/all' failed
解决:(http://blog.csdn.net/nicky_lyu/article/details/53236654)

7./usr/include/boost/python/detail/wrap_python.hpp:50:23: fatal error: pyconfig.h: 没有那个文件或目录
解决:(http://www.linuxdiyf.com/linux/26522.html)

8.
解决:(http://www.linuxidc.com/Linux/2015-04/116452.html)

9."error while loading shared libraries: xxx.so.x“”
解决:(http://blog.csdn.net/sahusoft/article/details/7388617)

10.recipe for target '.build_release/lib/libcaffe.so.1.0.0' failed
make: *** [.build_release/lib/libcaffe.so.1.0.0] Error 1
解决:(https://stackoverflow.com/questions/44067801/caffe-installation-linking-issue)

11.Ubuntu ldconfig(解决*.so不是符号连接)
解决:http://blog.csdn.net/hjxu2016/article/details/69389597

12.

Warning! ***HDF5 library version mismatched error***
The HDF5 header files used to compile this application do not match
variable 'HDF5_DISABLE_VERSION_CHECK' to a value of '1'.
Setting it to 2 or higher will suppress the warning messages totally.
Headers are 1.10.1, library is 1.8.16

解决:conda uninstall hdf5-1.10.1
conda install https://repo.continuum.io/pkgs/free/linux-64/hdf5-1.8.16-0.tar.bz2

13.import caffe失败 No module named caffe

解决:http://blog.csdn.net/u010417185/article/details/53559107

14.装了anaconda,和opencv,但import cv2报错:no module named cv2
解决:copy  opencv中 cv2.so 至anaconda2/lib/python2.7/site-packages内即可在python中使用

15.ssh安装
sudo apt-get install openssh-server

16.tensorflow gup 安装
pip install –ignore-installed –upgrade https://storage.googleapis.com/tensorflow/linux/gpu/tensorflow_gpu-0.12.0rc0-cp27-none-linux_x86_64.whl

17.keras安装
pip install keras 


18.install pycaffe interface
ImportError: No module named google.protobuf.internal
解决:http://www.cnblogs.com/darkknightzh/p/5993405.html
在protobuf-2.6.1的文件夹内,有python文件夹的README.txt里面写了怎么安装python的支持。

19.编译caffe出现错误:make: *** [.build_release/src/caffe/common.o] Error 1 
解决:http://blog.csdn.net/u011070171/article/details/52292680

20.CXX/LD -o .build_release/examples/cpp_classification/classification.bin
/usr/bin/ld: .build_release/examples/cpp_classification/classification.o: undefined reference to symbol '_ZN2cv6imreadERKNS_6StringEi'
//usr/local/lib/libopencv_imgcodecs.so.3.0: error adding symbols: DSO missing from command line
collect2: error: ld returned 1 exit status
Makefile:565: recipe for target '.build_release/examples/cpp_classification/classification.bin' failed
make: *** [.build_release/examples/cpp_classification/classification.bin] Error 1
解决:add the opencv_imgcodecs to the MakeFile 
LIBRARIES += opencv_core opencv_highgui opencv_imgproc opencv_imgcodecs










猜你喜欢

转载自blog.csdn.net/qq_33228039/article/details/78542772