Secured Banking Authentication by Pixel Correlation for Online Shopping

Abstract

This paper presents a secure way for bank transaction during online shopping with the help of graphical passwords that is image processing.The project's aim is to assist users in choosing and constructing a stronger password that cannot be easily compromised. This paper also talks about the clustering and recommendation of items that are bought by the user’s during online shopping. Item based recommendation is used so that unwanted items that are not relevant to the user’s shopping arenot recommended [Noise recommendation]. Multilevel clustering is used to cluster all the items and then group them. This project helps to avoid phishing attack.The attacker uses phishing e-mails to send different types of links or attachments that can be compromised, including credential details such as login details or account information of the user, in this type of cyber-attack.This method was introduced to overcome textual passwords, pins and other trivial passwords methods which were difficult to remember and prone to external attack. When a client forgets their password, they must create a new one that is both difficult to remember and easy to hack. A graphical password is easier to remember than a text password, and users can generate their own passwords using images. As a result, some researchers have proposed using a graphical password instead of a text-based one. As a result, several researchers have suggested graphical password as an alternative to text-based password. The user will select a particular part of the screen to confirm the process. We propose a framework having pixel correlation for user verification based on double hashing.