A Novel Graphical Password Authentication Scheme With Improved Usability

Abstract

Graphical password authentication schemes are gaining popularity as alternatives to traditional text-based passwords due to their potential to enhance security and usability. However, existing schemes often struggle with memorability and vulnerability to attacks. This project introduces a novel graphical password authentication scheme that combines image recognition and user-generated patterns to provide a multi-layered authentication process. Advanced image processing techniques are employed to prevent shoulder surfing and unauthorized access. A usability study was conducted, demonstrating improved memorability and user satisfaction, making this approach a promising solution for secure authentication.

Present System

In current graphical password systems:

• Users select images or patterns for authentication.
• Some systems use knowledge-based authentication, where users answer questions related to their images.
• Persuasion techniques may be used to encourage secure password choices.
• Security Issues: Vulnerable to shoulder surfing, brute-force attacks, and memorability challenges.

Disadvantages of Present System

1. Limited image choices increase predictability.
2. Users may struggle to remember complex patterns.
3. Knowledge-based authentication adds cognitive load.
4. Persuasive techniques may not always be effective, making passwords weaker.

Proposed System

The new graphical password authentication scheme aims to:

• Enhance security by combining image selection with user-generated patterns.
• Improve memorability through personalized images.
• Prevent attacks with advanced image processing techniques.
• Conduct a usability evaluation study with diverse participants to test effectiveness.

Advantages of Proposed System

• Enhanced security through multi-layered authentication.
• Better memorability through personal image selection.
• Improved security with multi-layered authentication.
• Resistance to attacks like shoulder surfing and brute-force attempts.
• User satisfaction due to intuitive password creation.
• Potential for wider adoption due to usability improvements.

Hardware Requirements

• Processor: Pentium IV 2.4 GHz
• Hard Disk: 40 GB
• RAM: 512 MB

Software Requirements

• Operating System: Windows
• Programming Language: Python
• Libraries: OpenCV (Image Processing), TensorFlow (ML), Pandas, Matplotlib

Output Screens

• User Signup Page: Users register with image and pattern selection.
• Login Page: User uploads the selected image and enters the correct pattern.
• Successful Login Screen: Displays welcome message if authentication is correct.
• Authentication Failure Screen: Shows error if incorrect image or pattern is entered.
• Admin Dashboard: View user authentication logs and security performance.