1. Introduction
Personnel security is becoming increasingly important in today's
modern world [1]. Biometric-based access control is one of the most
important technologies for cyber–physical security, and has received
increasing attention over the past two decades. In the competitive businessworld
of today, the need and demand for a biometric physical security
solution have never been higher. The biometricmarket is increasing
each year and this trend is set to continue, due to the increasing need for
security at borders, and in buildings, airports, etc. [2]. At its core, it aims
to identify a person with one or more of their body features, such as
their face, hand, fingerprint, or voice [1–3]. These biometric modalities
can be deployed for different applications including; searching for people,
remote access control, and secure corridors in airports.
To date, there has been a large amount of work done in biometrics,
with most of it focusing on using a single biometric mode. Recently,
the trend has been to build robust person identification systems based
on multimodal approaches, i.e., a combination of biometric features.
However, to obtain a robust multimodal solution, it is of benefit to employ
individual modalities which have good performance in isolation.
Furthermore, there is still much room for improvement with respect
to single mode approaches.
Human recognition through distinctive facial features supported by
an image database is still an appropriate subject of study as already
mentioned.We should not forget that this problemstill presents various
difficulties. For example, what will happen if an individual's haircut is
changed? Ismake-up a determining factor in the process of verification?
Would it significantly distort facial features? For these reasons, the
study of different parts of a face still merits investigation in order to
improve identification. Consequently, the analysis of lip contours is receiving
greater attention [4,5], as it is particularlywell-suited to deployment
on mobile phone platforms. The importance of lip features as
biometrics is reported in [6], where numerous lip-based features are
evaluated. Therefore in this work, an approach based on the shape of
lips is presented.
1.1. Related work
In this section, we briefly review work closely related to ours. Early
work in this area involved the tracking of lips, using features extracted
fromcolor distributions around the lip area [7]. The resulting feature dimensionality
was reduced using principal component analysis (PCA),
and classification was performed by linear discriminant analysis. By
combining this approach to lipmovement analysiswith speech analysis,
a significant improvement in speaker verification in noisy conditions