DB-ResNet:Real-time Scene Text Detection with Differentiable Binarization 问题总结 win10

配置https://github.com/MhLiao/DB
1、self._handle = _dlopen(self._name, mode) OSError: [WinError 126] 找不到指定的模块

D:\ProgramData\Anaconda3\envs\pytorch_gpu\python.exe E:/ZhuoZhuangOCR/Paper/Latest/DB-Resnet/DB/demo.py
Traceback (most recent call last):
  File "E:/ZhuoZhuangOCR/Paper/Latest/DB-Resnet/DB/demo.py", line 7, in <module>
    from experiment import Structure, Experiment
  File "E:\ZhuoZhuangOCR\Paper\Latest\DB-Resnet\DB\experiment.py", line 4, in <module>
    from structure.representers import *
  File "E:\ZhuoZhuangOCR\Paper\Latest\DB-Resnet\DB\structure\representers\__init__.py", line 1, in <module>
    from .seg_detector_representer import SegDetectorRepresenter
  File "E:\ZhuoZhuangOCR\Paper\Latest\DB-Resnet\DB\structure\representers\seg_detector_representer.py", line 3, in <module>
    from shapely.geometry import Polygon
  File "D:\ProgramData\Anaconda3\envs\pytorch_gpu\lib\site-packages\shapely\geometry\__init__.py", line 4, in <module>
    from .base import CAP_STYLE, JOIN_STYLE
  File "D:\ProgramData\Anaconda3\envs\pytorch_gpu\lib\site-packages\shapely\geometry\base.py", line 18, in <module>
    from shapely.coords import CoordinateSequence
  File "D:\ProgramData\Anaconda3\envs\pytorch_gpu\lib\site-packages\shapely\coords.py", line 8, in <module>
    from shapely.geos import lgeos
  File "D:\ProgramData\Anaconda3\envs\pytorch_gpu\lib\site-packages\shapely\geos.py", line 145, in <module>
    _lgeos = CDLL(os.path.join(sys.prefix, 'Library', 'bin', 'geos_c.dll'))
  File "D:\ProgramData\Anaconda3\envs\pytorch_gpu\lib\ctypes\__init__.py", line 364, in __init__
    self._handle = _dlopen(self._name, mode)
OSError: [WinError 126] 找不到指定的模块。

shapely版本有问题

conda install -c conda-forge shapely 

2、需要在本地电脑安装CUDA+cudnn,在anaconda安装不行,必须本地也安装对应的CUDA+cudnn版本,并多取一个变量名在这里插入图片描述

  echo $CUDA_HOME
  cd assets/ops/dcn/
  python setup.py build_ext --inplace

再运行python setup.py build_ext --inplace,才不会不会提示CUDA_HOME is not set。其原因如下:
DB-ResNet:Real-time Scene Text Detection with Differentiable Binarization 问题总结 win10_第1张图片

3、[未解决]FileExistsError: [WinError 183] 当文件已存在时,无法创建该文件。: ‘./outputs/workspace\DB’ -> 'workspace’

  File "E:\ZhuoZhuangOCR\Paper\Latest\DB-Resnet\DB\concern\log.py", line 68, in _make_storage
    os.symlink(storage_dir, self.log_dir)
FileExistsError: [WinError 183] 当文件已存在时,无法创建该文件。: './outputs/workspace\\DB' -> 'workspace'

4、将config中的yaml的路径改为绝对路径
在这里插入图片描述

你可能感兴趣的:(填坑之旅)