Summary
======
This article describes how certain types of captchas (such as the ones used 
by a German online-banking site) can be automatically recognized using 
software. The attack does not recognize one particular captcha itself but 
exploits a design error allowing to average multiple captchas containing 
the same information. The result can be recognized by conventional OCR 
programs thereby defeating the captcha. 

Details
======
The detailed article (including sample images) is online here:
http://www.cip.physik.uni-muenchen.de/~wwieser/misc/captcha/ 

Countermeasurements
==================
Website developers can easily defend against this attack by not 
allowing the extraction of a series of different captcha images 
with same content. Instead, the image should change only when the 
text content changes. 

Captcha designers can defend agaist averaging attacks by not using 
noise-like distortions. For example, moving and rotaing individual 
letters by a large enough distance/angle will spoil averaging by 
reducing the contrast in averaged images.

Contact: wwieser (at) gmx -dot- de
PLEASE do not CC me when posting to the list; I am subscribed.