University of Miami professor Mohamed Abdel-Mottaleb has developed a new way to improve the efficiency and accuracy of facial recognition technology. Abdel-Mottaleb and colleagues have developed a system that can use three-dimensional (3D) facial images, or combine two-dimensional images of the face with 3D models of the ear, to identify people by their unique facial features. The system uses fewer distinguishable landmarks of each face when matching 3D facial data, and is designed to automatically select the most discriminative facial regions, which are primarily within the regions of the nose, eyebrows, mouth, and chin.
The system also is designed to obtain a set of facial landmarks from frontal facial images and combine this data with a 3D ear recognition component. The researchers achieved a 100 percent identification rate when they combined the scores of these two modalities.
The team now wants to apply the methods to faces demonstrating facial expressions and to recognize faces using only profile images.
From University of Miami
View Full Article
Abstracts Copyright © 2009 Information Inc., Bethesda, Maryland, USA
No entries found