Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Histogram-based 3D face recognition

: Ullrich, Martin
: Zhou, Xuebing

Darmstadt, 2007, 74 S.
Darmstadt, Hochschule, Dipl.-Arb., 2007
Fraunhofer IGD ()
biometric; 3D face recognition; histogram

Biometric technology bears the potential to serve the ongoing demand for secure and reliable authentication, while at the same time providing unmatched ease of use. Among many different biometric modalities, face images are popular for being easy to capture and delivering very good recognition results. However, current applications commonly employ two-dimensional data, which have shown to be vulnerable to spoofing attacks and image quality problems. 3D face recognition is a promising technology that helps to overcome these issues and therefore has attracted a lot of attention recently.
This work is a research on histogram-based feature extraction for 3D face recognition. The proposed algorithm is based on simple statistic analysis of depth value distribution across the face surface and can be implemented very efficiently. The FRGC version 2 range image database has been used for testing to demonstrate the capabilities in connection with pose variation and facial expressions.