Open Access

ARTICLE

Distinction Between Real Faces and Photos by Analysis of Face Data

Byong Kwon Lee¹, Yang Sun Lee^2,*

1 Dept. of Multimedia Engineering, College of Engineering, Dongguk University, Seoul, 04620, South Korea
2 Div. of Convergence Computer & Media, Mokwon University, Daejeon, 35349, South Korea

* Corresponding Author: Yang Sun Lee,

Intelligent Automation & Soft Computing 2020, 26(1), 133-139. https://doi.org/10.31209/2019.100000134

Abstract

Biometric user authentication using the face has been applied mainly to access control systems. However, access is allowed even when a photo is presented instead of an actual face. This can facilitate illegal access including attending as a substitute or substitute authentication. An alternative approach has been implemented to solve this problem. The approach determines between a real face and a photo of a face using a UV sensor but this requires substantial cost and installation process because additional hardware (the UV sensor) is necessary. This paper proposes a three-step approach to identify between a real image and a photo. Step 1 determines authenticity using the background data and eliminating the face data. Step 2 determines authenticity using eyelid blinking on the face and facial gestures. Step 3 authorizes the user by extracting the feature points on the face.

---

Keywords

---