ABSTRACT
Usually, a unimodal biometric has its own limitations. For instance,
face recognition does not perform well with pose variations and
is sensitive to illumination and shadows, fingerprint recognition
cannot be applied for identification at a distance, and voice
recognition can be affected by age, flu, or throat infection. In order
to increase population coverage, extend the range of environmental
conditions, improve resilience to spoofing, and achieve higher recog-
nition accuracy, multimodal biometrics was introduced. Multimodal
biometrics use more than one means of physiological or behavioral
characteristics to identify and verify a person. In this chapter, the
advantages and limitations of each kind of biometrics technologies
are discussed. Traditional multimodal integration measures are
reviewed and analyzed, including integration in multiple traits,
integration in multiple snapshots of the same trait, and integration
in multiple representations and matching algorithms of the same
trait. The score fusion methods and applications in multimodal
biometrics are discussed in the chapter as well.