A Dynamic CAPTCHA Based on Persistence of Vision
Tao Men, Xiuqiong Zhang, Guo Huang, Yan Sun · 2012
With the rapid development of artificial intelligence and image recognition, traditional simple CAPTCHA is facing the challenge of being cracked, while too complicated CAPTCHA is difficult for the humans to recognize. After summarizing some kinds of commonly-used CAPTCHA, this paper proposes a novel dynamic CAPTCHA based on persistence of vision, which is formed by multi-frame images with useful characters and interfering characters. In each frame image, the useful characters are fixed and the interfering characters are changed constantly. When the multi-frame images are switched quickly, humans can get the useful characters because their eyes have the phenomenon of persistence of vision. So the CAPTCHA is easy for humans to recognize while hardly cracked by the computers. Theoretical analysis and simulation experiments illustrate that this scheme is effective, safe and with low complexity.