This paper concerns the conversion of color images to monochromatic form for the purpose of human face recognition. Many face recognition systems operate using monochromatic information alone even when color images are available. In such cases, simple color transformations are commonly used that are not optimal for the face recognition task. We present a framework for selecting the transformation from face imagery using one of three methods: Karhunen-Loève analysis, linear regression of color distribution, and a genetic algorithm. Experimental results are presented for both the well-known eigenface method and for extraction of Gabor-based face features to demonstrate the potential for improved overall system performance. Using a database of 280 images, our experiments using these methods resulted in performance improvements of approximately 4% to 14%.
This article is part of the series Biometric Signal Processing.
Optimization of Color Conversion for Face Recognition
1 Bradley Department of Electrical and Computer Engineering, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061-0111, USA
2 Department of Computer Science, Seattle Pacific University, Seattle, WA 98119-1957, USA
EURASIP Journal on Advances in Signal Processing 2004, 2004:948790 doi:10.1155/S1110865704401073
The electronic version of this article is the complete one and can be found online at: http://asp.eurasipjournals.com/content/2004/4/948790
|Received:||5 November 2002|
|Revisions received:||16 October 2003|
|Published:||21 April 2004|
© 2004 Jones and Abbott