一种三层动态验证码的构造方法研究  

A 3-layer Dynamic CAPTCHA Implementation

作　　者：崔竞松[1] 张五州[1] 彭洋[1] 梁玉[1] 肖邦[1] 梅婧婷[1] 张达[1] 汪霞[1] 

机构地区：[1]武汉大学计算机学院,武汉430072

出　　处：《武汉理工大学学报》2010年第11期122-125,共4页Journal of Wuhan University of Technology

基　　金：国家自然科学青年基金(60703009;60603012);国家自然科学基金(90718006);国家863计划项目(2008AA01Z404)

摘　　要：一种安全实用的三层动态验证码,创造性地将单帧零知识理论与生物视觉理论结合了起来,既保证了任一单帧都很难识别,又能确保人眼易于识别。充分利用了计算机在复杂背景下对多个运动物体进行识别的弱点,确保了计算机程序利用多帧仍很难破解。分层结构使得验证码的设计更为清晰,具有很强的扩展性和很大的优化空间。In order to prevent the attack from malicious progroms,the server system introduces CAPTCHA mechanism to tell humans and computers apart.With the continuous development of pattern recognition and artificial intelligence technology,the traditional 2D static CAPTCHA reveals more and more security vulnerabilities,making it possible that some malicious network programs can initiate the program attack by cracking the CAPTCHA.This paper presents a safe and practical 3-layer dynamic CAPTCHA,creatively combining the single-frame zero-knowledge theory and the biological vision theory,thus making it not only difficult to identify a single frame,but also easy to recognize for humans.This design takes advantage of the weakness of computers in identifying multiple moving objects in a complex context,ensuring it still extremely hard for the computer programs to crack even using multiple frames.Hierarchical structure makes the CAPTCHA design much more clearly,bearing strong scalability and large space for optimization.

关 键 词：验证码 三层 单帧零知识 运动物体识别 生物视觉理论 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]