The image (right) was created using software designed to generate pictures with hidden figures. Pattern-recognition software tends to have more trouble spotting the flamingo on the right than humans do. (Credit: Niloy J. Mitra, Hung-Kuo Chu, Tong-Yee Lee, Lior Wolf, Hezy Yeshurun, Daniel Cohen-Or)

The new technology uses simple images of a running man, or galloping horse, and converts them into blotches, hidden within a similarly blotchy scene. Computers are usually unable to detect the figure, but the human eye typically can.

"Captcha" or Completely Automated Public Turing test to tell Computers and Humans Apart traditionally generates distorted text or numbers, often often against a cluttered background. The user must types the correct string of characters in order to access an online service, such as the account creation tool for a free e-mail address.

But existing Captcha system, like a lot of other systems don't offer complete security and can be broken by security researchers and hackers. Captcha systems for Live Mail, Gmail, Yahoo!, Livejournal, and PayPal have all been cracked at some point.

So Captcha system designers must keep improving methods to stay that jump ahead of the bots.

"The systems we all use today are relatively easy to break," says Professor Danny Cohen-Or, (left)  a researcher on the project and a professor of computer science  at Tel Aviv University. "What we have developed is something that, with more effort, could be like the base of a stronger Captcha."

Developed with researchers at the Indian Institute of Technology in Delhi, the National Cheng Kung University in Taiwan, and others from the University of Tel Aviv, the software was inspired by the idea of the whole being greater than the sum of the parts.
Specifically, the software exploits the human ability to analyze a chaotic, fragmented scene to find a hidden figure.

The key was designing an adjustable system that could generate images easy enough for human identification, but too difficult for pattern-recognition software.

The software begins with a 3-D running dog, horse or man, converting the subject into a series of carefully generated black dots or "splats," that take into account the object silhouette and shape.

Longer, complicated shapes are broken into smaller parts, and silhouettes slightly deformed. Software then places the subject into a scene of more shapes, including some made of small pieces of the subject, to create added visual confusion. Videos are created as a series of still images.

The "emergent" images generated by the system tested on three kinds of learning-based pattern recognition software and after training on a set of 30 emergent images, the systems presented with newemergent images. The best of the three pattern recognition systems could only distinguish between a horse and a human 60 % of the time.

Humans, presented with the same task, answered correctly nearly 100% of the time. The software can also create images that are far more difficult for computers to interpret, but equally harder for humans to interpret, too.

Cartoon Courtesy