pva-faster-rcnn安装error汇总

总结一下，留备用

question1:

File"setup.py", line 10, in

    from setuptools import setup

ImportError:No module named setuptools

make:*** [all] Error 1

 

solution1:

wget http://pypi.python.org/packages/source/s/setuptools/setuptools-0.6c11.tar.gz  

tar zxvf setuptools-0.6c11.tar.gz  

cd setuptools-0.6c11  

python setup.py build  

python setup.py install

 

question2:    File "setup.py", line 12, in

    from Cython.Distutils import build_ext

ImportError:No module named Cython.Distutils

make:*** [all] Error 1

 

solution2:    

在Ubuntu14.04上，建议通过下面的方法安装，这是一种通用的方法，

手动下载下来就行了

wgethttps://bootstrap.pypa.io/get-pip.py --no-check-certificate

sudo python get-pip.py

sudo pip install cython

(add q2:

sudo python get-pip.py 

error:

pip isconfigured with locations that require TLS/SSL, however the ssl module inPython is not available.

Collecting pip

 Could not fetch URL https://pypi.python.org/simple/pip/: There was aproblem confirming the ssl certificate: Can't connect to HTTPS URL because theSSL module is not available. - skipping

 Could not find a version that satisfies the requirement pip (fromversions: )

No matching distribution foundfor pip )

(solution add2:

Sudo reboot 重新输入就可以，应该是后台有程序占用，kill也可以。

)

question3:

File "setup.py", line14, in

    import numpy as np

ImportError: No module namednumpy

make: *** [all] Error 1

solution3:

sudo apt-get installpython-numpy

 

question4:

x86_64-linux-gnu-gcc: error:utils/bbox.c: No such file or directory

x86_64-linux-gnu-gcc: fatalerror: no input files

compilation terminated.

error: command 'x86_64-linux-gnu-gcc'failed with exit status 4

make: *** [all] Error 1

 

solution4:

 

cd $FRCN_ROOT/lib/utils

cython bbox.pyx

 

question5:

x86_64-linux-gnu-gcc:error: nms/cpu_nms.c: No such file or directory

x86_64-linux-gnu-gcc:fatal error: no input files

compilationterminated.

error: command'x86_64-linux-gnu-gcc' failed with exit status 4

make: *** [all] Error1

 

solution5:

手动用cython生成cpu_nms.c

cd $FRCN_ROOT/lib/nms

cython cpu_nms.pyx

 

quetions6:

x86_64-linux-gnu-gcc:error: pycocotools/_mask.c: No such file or directory

x86_64-linux-gnu-gcc:fatal error: no input files

compilationterminated.

error: command'x86_64-linux-gnu-gcc' failed with exit status 4

make: *** [all] Error1

 

solution6:

手动用cython生成_mask.c

cd $FRCN_ROOT/lib/pycocotools

cython _mask.pyx

 

question7:

make: ***[.build_release/src/caffe/util/benchmark.o] Error 1

In file included from./include/caffe/util/device_alternate.hpp:40:0,

                 from./include/caffe/common.hpp:19,

                 from./include/caffe/blob.hpp:8,

                 from./include/caffe/layer.hpp:8,

                 fromsrc/caffe/layer_factory.cpp:8:

./include/caffe/util/cudnn.hpp:5:19:fatal error: cudnn.h: No such file or directory

 

solution7:

CuDNN是专门针对Deep Learning框架设计的一套GPU计算加速方案，目前支持的DL库包括Caffe，ConvNet, Torch7等。

CuDNN可以在官网免费获得，注册帐号后即可下载。官网没有找到安装说明，下载得到的压缩包内也没有Readme. 不过google一下就会找到许多说明。基本原理是把lib文件加入到系统能找到的lib文件夹里， 把头文件加到系统能找到的include文件夹里就可以。这里把他们加到CUDA的文件夹下（参考这里）

tar-xzvf cudnn-6.5-linux-R1.tgz

cd cudnn-6.5-linux-R1

sudocp lib* /usr/local/cuda/lib64/

sudo cp cudnn.h /usr/local/cuda/include/

http://www.unjeep.com/q/894288692.html

 

question8:

opencv编程中undefined reference to 'cv::dft':

./obj/local/armeabi/libopencv_imgproc.a(templmatch.cpp.o):In function `cv::crossCorr(cv::Mat const&,cv::Mat const&,cv::Mat&, cv::Size_, int,cv::Point_,double, int)':

templmatch.cpp:(.text+0x47e): undefined reference to`cv::getOptimalDFTSize(int)'

templmatch.cpp:(.text+0x494): undefined reference to `cv::getOptimalDFTSize(int)'

templmatch.cpp:(.text+0x92c): undefined reference to`cv::dft(cv::_InputArray const&,cv::_OutputArray const&, int, int)'

templmatch.cpp:(.text+0xfa4): undefined reference to`cv::dft(cv::_InputArray const&,cv::_OutputArray const&, int, int)'

templmatch.cpp:(.text+0x1006): undefined reference to`cv::mulSpectrums(cv::_InputArray const&,cv::_InputArray const&,cv::_OutputArray const&, int, bool)'

templmatch.cpp:(.text+0x102e): undefined reference to`cv::dft(cv::_InputArray const&,cv::_OutputArray const&, int, int)'

 

solution8:

add"opencv_imgcodecs" in Makefile.(LIBRARIES += glog gflags protobufleveldb snappy \
lmdb boost_system hdf5_hl hdf5 m \
opencv_core opencv_highgui opencv_imgproc opencv_imgcodecs)
If you input "make all",the problem is the same again.But if youdelete all the file in build(rm -rf ./build/*) before "make all"(Iuse make clean ),you will success.I just success

 

quesiton9:

AR-o .build_release/lib/libcaffe.a

LD-o .build_release/lib/libcaffe.so.1.0.0-rc3

/usr/bin/ld:cannot find -l -lboost_system

/usr/bin/ld:cannot find -lhdf5 -lopencv_core

collect2:error: ld returned 1 exit status

make:*** [.build_release/lib/libcaffe.so.1.0.0-rc3] Error 1

 

solution9:

1.系统缺乏对应的库文件；

2.版本不对应；

3.库文件的链接错误；

4.库文件路径设置问题。

 

       对应第一第二种情况，可以通过下载安装lib来解决，ubuntu大多数可以直接通过apt-get来安装：

 

apt-getinstall libxxx-dev

 

       一般遇到这种问题笔者第一时间也是会去检查系统是否已安装该lib或者是否已选择正确版本（只是习惯问题），如果还是不能解决问题，那么，引起错误的原因不是链接错误就是库文件路径问题了。

 

       通过find或者locate指令定位到链接文件，查看链接文件是否正确的指向了我们希望的lib，如果不是，用 ln -s */libxxx.so.x */libxxx.so 指令修改它。

 

 

 

Importcaffe时错误提示

1.错误提示：ImportError: No modulenamed skimage.io

解决方法：

直接使用终端安装：

pip install -U scikit-image

如果提示不识别   pip  命令，在Ubuntu14.04（64bit）下，使用下面命令安装pip包管理软件，也可以使用新立得软件包搜索“scikit-image”安装。

wgethttps://bootstrap.pypa.io/get-pip.py  --no-check-certificate

sudo python get-pip.py

 

question10：

ImportError: No module namedgoogle.protobuf.internal

提示错误可使用新立得软件包搜索“python-protobuf”安装。

然后使用import caffe测试接口是否调试成功。

 

Question11：

ImportError:No module named easydict

解决方法：

sudo pip install easydict  

 

question12:

错误描述: 当在$FRCN_ROOT/lib下make时出现

Traceback(most recent call last):
File “./demo.py”, line 18, in module
from fast_rcnn.test import im_detect
File “/home/lijiajun/py-faster-rcnn-blog/tools/../lib/fast_rcnn/test.py”, line17, in module
from fast_rcnn.nms_wrapper import nms
File“/home/lijiajun/py-faster-rcnn-blog/tools/../lib/fast_rcnn/nms_wrapper.py”,line 9, in module
from nms.gpu_nms import gpu_nms
ImportError: /home/lijiajun/py-faster-rcnn-blog/tools/../lib/nms/gpu_nms.so:undefined symbol: _nms

Solution 12:

编辑setup.py

cd $FRCN_ROOT/lib

vim setup.py

将gpu_nms.pyx改为gpu_nms.cpp

 

#before

 

Extension('nms.gpu_nms',

    ['nms/nms_kernel.cu', 'nms/gpu_nms.pyx'],

    ...

 

#after

 

Extension('nms.gpu_nms',

    ['nms/nms_kernel.cu', 'nms/gpu_nms.cpp'],

   

 

修改gpu_nms.c文件后缀为.cpp

cd $FRCN_ROOT/lib/nms

mv gpu_nms.c gpu_nms.cpp

rm gpu_nms.so