A SURVEY AND ANALYSIS OF CURRENT CAPTCHA APPROACHES

摘要

Computer programs are misusing Internet services designed for humans. A CAPTCHA, Completely Automated Public Turing test to tell Computers and Humans Apart, is a standard security mechanism to defend against such attacks. Two fundamental issues with CAPTCHAs are usability and robustness It is important for a CAPTCHA to be both legible for humans and strong against malicious computer programs. Recently, computer vision and pattern recognition algorithms have broken many well-known CAPTCHAs Lack of security and usability in CAPTCHAs designed to protect popular websites such as Gmail and Yahoo mail, with almost 500 million users in July 2011, would cause huge problems. Therefore, security researchers have become motivated to discover techniques to improve CAPTCHAs. Exploiting the gap in the recognition abilities between humans and computers is a key point to design a CAPTCHA that is hard-to-break for machines but easy-to-solve for humans In this paper, we introduce current CAPTCHAs and attacks against them; we investigate the robustness and usability of current CAPTCHAs and discuss ideas to develop more robust and usable CAPTCHAs