Detecting the Orientation of N-gonal Cropped Sub-images and Its Application

An increasing number of public web services have attempted to prevent exploitation by bots and automated scripts, by requiring a user to solve a Turing-test problem, namely a ”Completely Automated Public Turing test to tell Cmputers and Humans Apart (CAPTCHA)”, before they are allowed to use web services. In this paper, we present an effective image-based CAPTCHA based on the orientation of N-gonal cropped sub-images as a solution of CAPTCHAs. Our CAPTCHA is based on the difficulty of detecting the orientation of N-gonal sub-images. In our CAPTCHA, the number of orientations and the crop size are important considerations, since our CAPTCHA requires users to find the orientation of sub-images cropped in the form of a regular polygon. So, we discuss usability of our CAPTCHA and the efficient values of the number of orientations and the crop size through user experiments and SVM-based machine learning tests in this paper.