Reputation: 37
I'm working on an image to text OCR system and I kept receiving an error on the console stating
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.
The score_test.py is stored in ../OCR/
score_test.py Code:
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?
Upvotes: 1
Views: 2241
Reputation: 4445
.DS_Store
is a "hidden" file in directories on Mac OS X.
In your if img != "Thumbs.db"
check, you should add:
if not img.startswith('.') and img != 'Thumbs.db':
this way you filter out all "hidden" files.
Upvotes: 4