An XML transformed method to improve effectiveness of graphical password authentication

Abstract

Graphical passwords are already examined as the reliable authentication scheme over the textual passwords. These passwords promise the better robustness and memorability. Various available applications use the image and graphical passwords based authentication systems. But, such authentication systems suffer from the problem of managing pictorial or image data. To map the user input image or graphical password over the larger database, slow down the authentication process. In this present work, XML based schema is proposed to represent the graphical image. As a password image with the graphical pattern is loaded by the user, the server processed the pattern and verifies it for valid pattern based on stroke length and drift. Various forms of graphical patterns are obtained on input graphical pattern by applying the numerous transformations. The extracted pixel values from these practices are kept in an XML pattern DB. Then the server also updated the pattern bits in input image using LSB steganography approach and returns it back to the user as password image. Each time, user input the password image, the password pattern extraction is performed, and mapping is done on XML pattern DB. The implementation of the presented model is done as Mobile and Desktop application. The comparative performance evaluation signifies that the method is more effective than other image password mappers. As the complete information is extracted from the query password pattern image, the 100% accuracy in password mapping is obtained. The applied qualitative measures also verify that the proposed model has improved the reliability and robustness against various aspects.