Author(s): Prashant Kumar Abhishek , Savita Shiwani .

Published in: International Journal of Innovative Science and Research Technology


Volume/Issue: Volume 1 Issue 7

Google Scholar: :-

Scribd :-

In recent years, a considerable number of public services in the web have been trying to prevent abuse from automated programs by requiring from the users the resolution of a challenge in the format of a “Turing test” [12], popularly known as CAPTCHA, “Completely Automated Public Turing test to tell Computers and Humans Apart”. In these services, the users are only able to start using the service after providing a successful answer to the test. The main advantage of leveraging such tests is that, theoretically, the tests are easily generated and answered by humans. All CAPTCHAs possess some sort of secret information which is initially only known by the challenger but not by the agent being challenged. In this Paper, we are working for improve the performance of the design Captcha so that security can be enhance. In the previous paper [18], they are working for the Captcha protocol. In which Captcha is design by the graphical representation. User has to see the Captcha and type the same text into the given box. We are using unsupervised Captcha which divide the given input image in Auxillary visual word, Bow model, Affinity matrix, image annotation. By this process the execution time will get reduce and security of the Captcha will get improve.
Keyword:- CARP, E-mail attacks, Online guessing attacks, Relay attacks, Shoulder surfing attack.