Add hust_code OCR

README.md

@@ -9,11 +9,13 @@ Just modify config.py and enjoy!
 ## requirements
 
 ```bash
-sudo pacman -S python-selenium geckodriver
+sudo pacman -S python-selenium geckodriver python-pillow tesseract tesseract-data-eng
 ```
 
 Any linux with GUI support. it launches your firefox.
 
+**NOTE:** Run `tesseract example.png out.txt` and MAKE SURE it works! The result MUST be `8518`.
+
 ### WARNING: While making this script, I assume that your system language is ENGLISH
 
 一个不必要的提示：如果你在墙内，用recolic.org代替recolic.net。不然CCP会阻止你的访问。

hust_code.py

+from PIL import Image
+import sys
+
+i, o = sys.argv[1:]
+
+im = Image.open(i)
+# In pass.hust.edu.cn verification code, the second frame is COMPLETE!
+# let's get the second frame.
+im.seek(im.tell()+1)
+im.save(o)
+
+im = Image.open(o)
+# To white-black
+im = im.convert('L')
+im = im.point(lambda x: 0 if x<254 else 255, '1')
+# Try to crop
+w, h = im.size
+if w == 90 and h == 58:
+    im = im.crop((0, 18, w, 39))
+else:
+    print('WARNING: Expected image w=90, h=58, Got image w={}, h={}.'.format(w, h))
+# Save for OCR
+im.save(o)
+
+
+