This paper is published in Volume-11, Issue-3, 2025
Area
Image Processing
Author
Abhishekayya Kambi, Ankit Ronad, Sumanth Mudegoudra, Dr Vidyagouri B
Org/Univ
SDM College of Engineering and Technology, Dharwad, India
Pub. Date
27 May, 2025
Paper ID
V11I3-1203
Publisher
Keywords
Spoofing, Facial Landmarks, Texture, Motion Blur, Eye-Blink

Citationsacebook

IEEE
Abhishekayya Kambi, Ankit Ronad, Sumanth Mudegoudra, Dr Vidyagouri B. Anti-Face Spoofing Detection using Texture and Eye Blink Parameters, International Journal of Advance Research, Ideas and Innovations in Technology, www.IJARIIT.com.

APA
Abhishekayya Kambi, Ankit Ronad, Sumanth Mudegoudra, Dr Vidyagouri B (2025). Anti-Face Spoofing Detection using Texture and Eye Blink Parameters. International Journal of Advance Research, Ideas and Innovations in Technology, 11(3) www.IJARIIT.com.

MLA
Abhishekayya Kambi, Ankit Ronad, Sumanth Mudegoudra, Dr Vidyagouri B. "Anti-Face Spoofing Detection using Texture and Eye Blink Parameters." International Journal of Advance Research, Ideas and Innovations in Technology 11.3 (2025). www.IJARIIT.com.

Abstract

Growing reliance on facial recognition for secure authentication in various applications, ensuring that facial inputs are genuine and not spoofed using photos, videos, or masks has become critical. This work introduces a real-time anti-face spoofing detection system that harnesses computer vision and deep learning to verify the liveness of facial inputs. The system integrates Media Pipe Face Mesh for accurate facial landmark detection, a Convolutional Neural Network (CNN) for classifying real vs fake faces, and eye blink detection using Eye Aspect Ratio (EAR) to further enhance liveness verification. Additionally, a texture analysis module and motion blur detection help assess image quality and prevent spoofing attempts through printed photos or video replays. A dynamic overlay displays relevant metrics such as EAR, texture score, model confidence, and blur score, aiding both real-time feedback and system transparency. The interface includes a timestamp module and real-time performance chart for enhanced monitoring. This robust solution contributes to secure biometric authentication by combining multiple detection layers for high accuracy in face liveness classification.