Race Classification using Craniofacial Features from Colored Face Images
Abdul-Husain M. Abdullah1, Enas W. Abood2
1Abdul-husain M. Abdullah, Department of Computer, University of Basrah, College of Science, Basrah , Iraq.
2Enas W. Abood, Department of Computer, University of Basrah, College of Science, Basrah , Iraq.
Manuscript received on July 26, 2014. | Revised Manuscript received on August 04, 2014. | Manuscript published on August 30, 2014. | PP: 138-143 | Volume-3 Issue-6, August 2014. | Retrieval Number: F3349083614/2013©BEIESP
Open Access | Ethics and Policies | Cite
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: This paper produces a system for race classification from face images. Two powerful types of local features have been considered: craniofacial features (eyes, mouth, nose) of the faces and color variance of the skin color together to further improves race classification accuracy. For classification, five ratios have been taken which calculated as a mathematically relation between features using four race groups selected from FG-NET ,CPIR database and other gathered by us as own database. The system scored a success about 82% in recognition for tested images.
Keywords: Race recognition, Facial features.