Abstract: Users who enter their credentials in a public location risk having their passwords stolen by intruders. An attacker can steal a password by watching the individual's authentication session or recording it. Shoulder-surfing is a well-known risk, and it's especially dangerous while authenticating in public. Until recently, the user's awareness was the sole barrier against shoulder-surfing. Shoulder surfing resistant password authentication ensures that the user's authentication is not compromised by shoulder surfing.Because the user is never required to click directly on password symbols, it allows users to authenticate by inputting their password in a graphical manner in vulnerable areas. This mechanism's usability testing revealed that inexperienced users were able to correctly input and remember their graphical password With everything turning online, the risk of cybercrimes and privacy breaches is also increasing. Passwords play a huge role in keeping your data safe online as well as offline platforms. Passwords are the default method of authentication to get access to our accounts. There are various types of authentication available for users to secure their accounts.

Keyword: Image Selection ,Extraction , Machine learning

| DOI: 10.17148/IARJSET.2022.94112