Human Face Recognition Technology Using the Karhunen-Loeve Expansion Technique

Authors

Anthony Giordano, Regis UniversityFollow
Michael Uhrig, Regis University

Abstract

We will explore the area of face recognition using the partial singular value decomposition of a matrix and test some of its successes and limitations. We constructed a database consisting of 130 pictures of 65 individuals, and then used the Karhunen-Loéve (KL) Expansion method to relate pictures from outside the database to those in the database. While this method was generally very successful, we were able to test and define several of its limitations.

Author Bio

Anthony Giordano is currently an undergraduate junior at Regis University, and is pursuing a double major in chemistry and mathematics. He truly enjoyed the opportunity to perform research at Regis University during the summer of 2005. Eventually, Anthony would like to pursue a doctorate in physical/organic chemistry, and work in a research environment where chemistry and mathematics meld together.

Michael Uhrig will be graduating in May 2006 from Regis University with a major in mathematics and a minor in philosophy. He plans to attend graduate school in mathematics.

Faculty Sponsor

Jim Seibert

Recommended Citation

Giordano, Anthony and Uhrig, Michael (2006) "Human Face Recognition Technology Using the Karhunen-Loeve Expansion Technique," Rose-Hulman Undergraduate Mathematics Journal: Vol. 7: Iss. 1, Article 11.
Available at: https://scholar.rose-hulman.edu/rhumj/vol7/iss1/11