Security analysis method of recognition-based graphical password

Abstract

One of the most important primitive security mechanisms is the authentication system. Authentication through the use of password is a commonly utilized mechanism for authentication of users. In general, users utilize characters as their password; however, passwords based on texts are hard to recall and if the passwords are too simple and predictable, then there is the danger of being susceptible to threats. In order to overcome the problems with authentication, an alternative and new approach has been introduced utilizing images for passwords. The idea gains support from the knowledge that the human's brain is highly capable of remembering many detailed images, however remembering texts are more difficult. Users who utilize the graphic authentication carry out certain functions on the images such as to click, drag, and movement of the mouse and so on. This research reviews several common Recognition-Based graphical password methods and analyzes their security based on the estimation criteria. Moreover, the research defines a metric that would make it possible for the analysis of the security level of the graphical passwords that are Recognition-Based. Finally, a table comparing the limits of each method based on the security level is presented.