Python - IO:错误无法识别图像文件

I'm working on an image to text OCR system and I kept receiving an error on the console stating

我正在处理一个图像到文本OCR系统，我不断收到控制台声明的错误

Traceback (most recent call last):
  File "crack_test.py", line 48, in <module>
    temp.append(buildvector(Image.open("./iconset/%s/%s"%(letter,img))))
  File "/Users/seng_kin/anaconda/lib/python2.7/site-packages/PIL/Image.py", line 2290, in open
    % (filename if filename else fp))
IOError: cannot identify image file './characterset/a/.DS_Store'

I have a folder in a ../OCR/characterset/ directory which has 26 subfolders representing 26 characters whereby each subfolder contains a .PNG of a character.

我有一个文件夹在…/OCR/字符集/目录，其中有26个子文件夹，表示每个子文件夹包含一个字符的. png。

The score_test.py is stored in ../OCR/

score_test。py存储在./OCR/中。

score_test.py Code:

score_test。py代码:

from PIL import Image
import os, sys
import math

****def functions****

iconset = ['a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r',
           's','t','u','v','w','x','y','z']

imageset = []

for letter in iconset:
    for img in os.listdir('./characterset/%s/'%(letter)):
        temp = []

        if img != "Thumbs.db": #windows check...
            temp.append(buildvector(Image.open("./characterset/%s/%s"%(letter, img))))

        imageset.append({letter:temp})

I saw other solutions on the website but all is related to the absence of from PIL import Image. Am I missing something?

我在网站上看到了其他的解决方案，但都与PIL import Image的缺失有关。我遗漏了什么东西?

1 个解决方案

#1

.DS_Store is a "hidden" file in directories on Mac OS X.

. ds_store是Mac OS X目录中的“隐藏”文件。

In your if img != "Thumbs.db" check, you should add:

在你的if img != "拇指"。db" check，你应该加上:

if not img.startswith('.') and img != 'Thumbs.db':

this way you filter out all "hidden" files.

这样可以过滤掉所有“隐藏”文件。

#1