3D Face Recognition

In our 3D face projects, we use 3D cameras, which are based on stereo and structured light, in order to create a 3D 'point cloud' on the face, which is then meshed. Simultaneously a standard 2D colour-texture image is captured, which is automatically registered with the 3D data.

We explore research carried out in the field of face recognition, with the ultimate aim of producing a highly effective face recognition algorithm, for use in such application areas as secure site access, suspect identification and surveillance.  A new line of research is proposed to analyse and compare the advantages offered by the various 2D (intensity image) approaches and newly emerging 3D (geometrical surface structure) approaches.  We develop and test both 2D and 3D face recognition systems on a large database of subjects and demonstrate how simple image pre-processing methods can significantly improve performance of existing 2D approaches.  Current results gathered from tests of our 3D face recognition system indicate that the approach is capable of achieving significantly lower error rates than existing 2D systems.

Face Recognition as a biometric.

Face Detection.

The Eigenface based method of Face Recognition.

3D Face Recognition.

3D Face Recognition - Graph Matcher (login required).

Current research also includes the following (information regarding which, will become available shortly):

•     Facial Feature Localisation.
•     The use of AURA in 3D facial graph matching.

Papers and other documents

 Face Recognition: A Literature Review
 Research area literature review in PowerPoint (1.61MB)

 Evaluation of Image Pre-processing Techniques for Eigenface Based Face Recognition.pdf
 A research paper evaluating the improvements gained by use of various image pre-processing techniques, when applied to the Eigenface based method of face recognition. (413KB)

 Evaluation of Image Pre-processing Techniques for Eigenface Based Face Recognition.ppt
 A short PowerPoint presentation (1.25MB)

 Face Recognition: A Comparison of Appearance-Based Approaches.pdf
 A research paper presented at DICTA 2003, comparing three methods of face recognition, namely the eigenface, fisherface and direct correlation methods, evaluating the improvements introduced to each method by the use of image pre-processing techniques (554KB)

 Face Recognition: A Comparison of Appearance-Based Approaches.ppt
 A short PowerPoint presentation (1.53MB)

 Combining multiple face recognition systems using Fisher’s linear discriminant.ps
 A research paper presented at the SPIE Defense and Security Symposium 2004, describing a method of utilising multiple 2D face spaces, with the discriminant as criteria for creating a more effective face recognition system (817KB)

 Three-Dimensional Face Recognition: An Eigensurface Approach.pdf
 A research paper describing the application of PCA techniques to three dimensional face data for recognition, evaluating the effectiveness of a range of surface representations (224KB)

 Three-Dimensional Face Recognition: A Fishersurface Approach.pdf
 A research paper presented at ICIAR 2004 describing the use of PCA combined with LDA techniques, as applied in the fisherface method of face recognition, to three dimensional face data, evaluating the effectiveness of a range of surface representations and distance metrics when compared with the eigensurface method (487KB)

 Three-Dimensional Face Recognition Using Surface Space Combinations.pdf
 A research paper presented at the BMVC 2004 describing the state-of-the-art in three dimensional face recognition, combining multiple surface space representations of 3D face data, in order to create a more effective face recognition system (342KB)

 Three-Dimensional Face Recognition Using Surface Space Combinations.ppt
 A short PowerPoint presentation (3.04MB)

Link to Tom Heseltine's PhD work.