Faster R-CNN CPU环境搭建
操作系统:
bigtop@bigtop-SdcOS-Hypervisor:~/py-faster-rcnn/tools$ cat /etc/issue Ubuntu 14.04.2 LTS \n \l
Python版本:
bigtop@bigtop-SdcOS-Hypervisor:~/py-faster-rcnn/tools$ python --version Python 2.7.6
pip版本:
bigtop@bigtop-SdcOS-Hypervisor:~/py-faster-rcnn/tools$ pip --version pip 1.5.4 from /usr/lib/python2.7/dist-packages (python 2.7)
环境变量情况:
bigtop@bigtop-SdcOS-Hypervisor:~/py-faster-rcnn/caffe-fast-rcnn$ echo $LD_LIBRARY_PATH bigtop@bigtop-SdcOS-Hypervisor:~/py-faster-rcnn/caffe-fast-rcnn$ echo $PATH /usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/usr/local/games
~/.bashrc内容,可以看到所有和PATH以及LD_LIBRARY_PATH相关的内容都没有设置:
# ~/.bashrc: executed by bash(1) for non-login shells. # added by Anaconda2 4.0.0 installer #export PATH="/home/bigtop/anaconda2/bin:$PATH" #export LD_LIBRARY_PATH="/home/bigtop/anaconda2/lib/":$LD_LIBRARY_PATH #export LD_LIBRARY_PATH="/lib/x86_64-linux-gnu/":$LD_LIBRARY_PATH
1. 安装Caffe需要的依赖包:
sudo apt-get install build-essential # basic requirement sudo apt-get install libprotobuf-dev libleveldb-dev libsnappy-dev libopencv-dev libboost-all-dev libhdf5-serial-dev libgflags-dev libgoogle-glog-dev liblmdb-dev protobuf-compiler #required by caffe
使用完上面的命令后,依赖已经安装完毕,但是由于Ubuntu 14.04版本的原因,导致opencv相关的环境不能够正常的work。所以,我重新编译了一个OpenCV,版本为3.1.0。
在解压后的目录中执行:
bigtop@bigtop-SdcOS-Hypervisor:~/tools/opencv-3.1.0$ cmake -DBUILD_TIFF=ON
然后执行make 和make install
2. 编译cafe-fast-rcnn
bigtop@bigtop-SdcOS-Hypervisor:~/py-faster-rcnn/caffe-fast-rcnn$ ls build CMakeLists.txt data examples LICENSE Makefile.config~ models scripts caffe.cloc CONTRIBUTING.md distribute include Makefile Makefile.config.example python src cmake CONTRIBUTORS.md docs INSTALL.md Makefile.config matlab README.md tools bigtop@bigtop-SdcOS-Hypervisor:~/py-faster-rcnn/caffe-fast-rcnn$ pwd /home/bigtop/py-faster-rcnn/caffe-fast-rcnn
修改这个目录下的Makefile.config(如果没有这个文件,就直接cp Makefile.config.example Makefile.config)
将CPU_ONLY := 1开关和WITH_PYTHON_LAYER开关打开:
然后在该目录下执行:make –j8 && make pycaffe
在此过程中,可能会出现各种和python相关的包缺失问题,这里记录下,以便查询:
A》将caffe-fast-rcnn/python目录下的requirements下的依赖都装一遍:
bigtop@bigtop-SdcOS-Hypervisor:~/py-faster-rcnn/caffe-fast-rcnn/python$ cat requirements.txt Cython>=0.19.2 numpy>=1.7.1 scipy>=0.13.2 scikit-image>=0.9.3 matplotlib>=1.3.1 ipython>=3.0.0 h5py>=2.2.0 leveldb>=0.191 networkx>=1.8.1 nose>=1.3.0 pandas>=0.12.0 python-dateutil>=1.4,<2 protobuf>=2.5.0 python-gflags>=2.0 pyyaml>=3.10 Pillow>=2.3.0 six>=1.1.0
执行如下命令:
for req in $(cat requirements.txt); do pip install $req; done
这里有一个小技巧,因为pip这个工具对应的网络非常的烂:
这个时候,可以将其改为国内的镜像网站,速度将提升几个数量级,方法如下:
新建~/.pip/pip.confg文件,内容如下:
[global] index-url = http://pypi.douban.com/simple
trusted-host = pypi.douban.com
或者在安装一个软件包的时候使用-i选项:
在我安装requirements.txt中涉及的依赖包的过程中,发现matplotlib始终没有安装成功,最后采用apt-get的方式进行了安装,如下:
sudo apt-get install python-matplotlib
B>opencv环境和caffe-fast-rcnn默认的Makefile配置有点小问题,cv::imread(cv:: String const&, int)找不到:
解决方案:
在一切都正常的情况下,对caffe-fast-rcnn进行make和make pycaffe的结果如下:
编译好caffe-fast-rcnn后,在py-faster-rcnn/lib中执行make命令:
bigtop@bigtop-SdcOS-Hypervisor:~/py-faster-rcnn/lib$ make python setup.py build_ext --inplace running build_ext skipping 'utils/bbox.c' Cython extension (up-to-date) skipping 'nms/cpu_nms.c' Cython extension (up-to-date) skipping 'pycocotools/_mask.c' Cython extension (up-to-date) rm -rf build bigtop@bigtop-SdcOS-Hypervisor:~/py-faster-rcnn/lib$
3. 在安装配置好caffe-fast-rcnn后,修改py-faster-rcnn相关配置,让其模型可以在没有GPU的环境下运行:
A>将 ~/py-faster-rcnn/lib/fast_rcnn/config.py的如下内容:
B>将 ~/py-faster-rcnn/tools/test_net.py和 ~/py-faster-rcnn/tools/train_net.py的caffe.set_mode_gpu()修改为caffe.set_mode_cpu().
C>将~/py-faster-rcnn/lib/setup.py中,含有'nms.gpu_nms’的部分去掉,去掉后的内容如下:
112 ext_modules = [ 113 Extension( 114 "utils.cython_bbox", 115 ["utils/bbox.pyx"], 116 extra_compile_args={'gcc': ["-Wno-cpp", "-Wno-unused-function"]}, 117 include_dirs = [numpy_include] 118 ), 119 Extension( 120 "nms.cpu_nms", 121 ["nms/cpu_nms.pyx"], 122 extra_compile_args={'gcc': ["-Wno-cpp", "-Wno-unused-function"]}, 123 include_dirs = [numpy_include] 124 ), 125 Extension( 126 'pycocotools._mask', 127 sources=['pycocotools/maskApi.c', 'pycocotools/_mask.pyx'], 128 include_dirs = [numpy_include, 'pycocotools'], 129 extra_compile_args={ 130 'gcc': ['-Wno-cpp', '-Wno-unused-function', '-std=c99']}, 131 ), 132 ]
D>做到上面三部后,还是不够的,还需要将:
../lib/fast_rcnn/nms_wrapper.py:9:#from nms.gpu_nms import gpu_nms
注释掉:
否则,会抛出如下的异常:
Traceback (most recent call last): File "./demo.py", line 18, in from fast_rcnn.test import im_detect File ".../py-faster-rcnn-master/tools/../lib/fast_rcnn/test.py", line 17, in from fast_rcnn.nms_wrapper import nms File ".../py-faster-rcnn-master/tools/../lib/fast_rcnn/nms_wrapper.py", line 11, in from nms.gpu_nms import gpu_nms ImportError: No module named gpu_nms
4. 运行demo.py
在环境一切就绪的情况下,将faster的模型下载下来:
bigtop@bigtop-SdcOS-Hypervisor:~/py-faster-rcnn/data/scripts$ ls fetch_faster_rcnn_models.sh fetch_imagenet_models.sh fetch_selective_search_data.sh
运行其中的./fetch_faster_rcnn_models.sh脚本就可以下载下来了。
在/home/bigtop/py-faster-rcnn/tools目录下运行, python demo.py --cpu:
最后的结果如下:
注意:由于我是在没有图形界面终端上运行的,默认情况下demo.py会假设运行在有图形界面的环境中,需要修改demo.py的地方如下:
首先,在demo.py代码的最前面,注意一定是最前面,否则可能不成功,加入如下两行:
其次,在plt.draw()的地方加入savefig()语句,将结果保存成jpg文件形式:
5. 其他在安装过程中遇到的问题(比较杂,记录于此),如果上面的四个步骤进行的比较顺利的话,是不会遇到下面这些问题的:
5.1 No module named skimage.io:
5.2 下面这个问题是因为缺少,easydict,使用 sudo pip install easydict可以解决:
bigtop@bigtop-SdcOS-Hypervisor:~/py-faster-rcnn/tools$ python demo.py --cpu Traceback (most recent call last): File "demo.py", line 17, in <module> from fast_rcnn.config import cfg File "/home/bigtop/py-faster-rcnn/tools/../lib/fast_rcnn/config.py", line 23, in <module> from easydict import EasyDict as edict ImportError: No module named easydict bigtop@bigtop-SdcOS-Hypervisor:~/py-faster-rcnn/tools$ sudo pip install easydict Downloading/unpacking easydict Downloading easydict-1.6.zip Running setup.py (path:/tmp/pip_build_root/easydict/setup.py) egg_info for package easydict Installing collected packages: easydict Running setup.py install for easydict Could not find .egg-info directory in install record for easydict Successfully installed easydict Cleaning up... sudo pip install easydict
5.3 这个问题是因为scipy安装出现问题,将其删掉:rm -fr /tmp/pip_build_root/scipy/,然后重新安装可以解决:
d --compile failed with error code 1 in /tmp/pip_build_root/scipy Traceback (most recent call last): File "/usr/bin/pip", line 9, in <module> load_entry_point('pip==1.5.4', 'console_scripts', 'pip')() File "/usr/lib/python2.7/dist-packages/pip/__init__.py", line 235, in main return command.main(cmd_args) File "/usr/lib/python2.7/dist-packages/pip/basecommand.py", line 161, in main text = '\n'.join(complete_log) UnicodeDecodeError: 'ascii' codec can't decode byte 0xe2 in position 98: ordinal not in range(128)
网上有很多搭建caffe的教程,都提到了用Anaconda,本来这个包是很好的,它可以解决很多python依赖的问题,可惜的是,它和我用的Ubuntu版本兼容性出现了问题,所以,我最终放弃了Anaconda,所有的python依赖都通过pip或者是apt-get进行了安装。
5.4 报下面这个错误,是因为caffe的环境都没有准备好,很有可能是没有执行make pycaffe:
Traceback (most recent call last): File "detector.py", line 29, in <module> import caffe File "/python/caffe/__init__.py", line 1, in <module> from .pycaffe import Net File "/caffe/pycaffe.py", line 6, in <module> from ._caffe import CaffeNet ImportError: No module named _caffe
5.5 error: undefined reference to `TIFFIsTiled@LIBTIFF_4.0'
error: undefined reference to `TIFFIsTiled@LIBTIFF_4.0'
这个就是上文中提到的,使用ubuntu自带的opencv库会出现的问题,解决办法就是重新编译opencv。
6. 总结
整个过程还是颇费周折,因为caffe依赖的东西太多,环境搭建费事费力,最好的办法还是弄一个docker镜像,这样才能够从环境搭建的苦海中解脱,可惜的是我从daocloud上down下来的镜像是不能够运行在cpu上的。
参考文档:
1. http://caffe.berkeleyvision.org/installation.html caffe官方文档
2. https://github.com/BVLC/caffe/issues/1276 cv::imread(cv::String const&, int)' collect2: error
3.https://github.com/BVLC/caffe/issues/263 can't import _caffe module
4.https://github.com/rbgirshick/py-faster-rcnn/issues/8 ImportError: No module named gpu_nms
5.http://www.cnblogs.com/empty16/p/4828476.html caffe环境搭建文章
6.https://github.com/BVLC/caffe/issues/50 make pycaffe error
7.http://blog.csdn.net/tangwei2014/article/details/45442275 Fast RCNN Ubuntu安装笔记
8.http://topmanopensource.iteye.com/blog/2004853 pip镜像加速
9. https://github.com/BVLC/caffe/issues/1276 ../lib/libcaffe.so: undefined reference to cv::imread
10. https://groups.google.com/forum/#!topic/caffe-users/wKYe45FKSqE Installation error: undefined reference to `TIFFIsTiled@LIBTIFF_4.0'
11. https://groups.google.com/forum/#!topic/caffe-users/0PrZlro7QbU undefined reference to `lzma_index_buffer_decode@XZ_5.0