A Fast Multi-Modal Approach to Facial Feature Detection

                              Chris Boehnen                                                                                 Trina Russ                         
           University of Notre Dame, cboehnen@nd.edu         Sandia National Laboratories, (tdruss@sandia.gov)

Sandia National Laboratories, cbboehn@sandia.gov                                                                                                       

Abstract
 

As interest in 3D face recognition increases the importance of the initial alignment problem does as well.  In this paper we present a method utilizing the registered 2D color and range image of a face to automatically identify the eyes, nose, and mouth.  These features are important to initially align faces in both standard 2D and 3D face recognition algorithms.  For our algorithm to run as fast as possible, we focus on the 2D color information.  This allows the algorithm to run in approximately 4 seconds on a 640X480 image with registered range data.  A 1,500 image test set achieved a facial feature detection rate of 99.6% with 0.4% of the images skipped due to hair obstruction of the face. 
 

Download this Article (pdf)

Keywords: Biometrics Facial Feature Detection 2D 3D Scanner Eyes Nose Mouth Color Map Confidence