INSTITUTE OF ROAD AND TRANSPORT TECHNOLOGY

ERODE-638316

FACE RECOGNITION
(pattern matching & bio-metrics)

CATEGORY- IMAGE PROCESSING

Submitted by,

G.MEGALA
R.KIRUTHIGA
Third year-CSE
Email: megala.deva95@gmail.com
Kirutirk95@gmail.com

Phone no:9659694536
9095471228
ABSTRACT: people's faces as they enter a defined area.
There is no intrusion or delay, and in most
Government agencies are investing a cases the subjects are entirely unaware of
considerable amount of resources into the process. They do not feel "under
improving security systems as result of surveillance" or that their privacy has been
recent terrorist events that dangerously invaded.
exposed flaws and weaknesses in today’s
safety mechanisms. Badge or password-
based authentication procedures are too
easy to hack. Biometrics represents a valid HISTORY:
alternative but they suffer of drawbacks as
well. Iris scanning, for example, is very Humans
reliable but too intrusive; fingerprints are have
socially accepted, but not applicable to always
non-consentient people. On the other hand, had the innate ability to recognize and
face recognition represents a good distinguish between faces, yet computers
compromise between what’s socially only recently have shown the same ability.
acceptable and what’s reliable, even when In the mid 1960s, scientists began work on
operating under controlled conditions. In using the computer to recognize human
last decade, many algorithms based on faces. Since then, facial recognition
linear/nonlinear methods, neural networks, software has come a long way.
wavelets, etc. have been proposed.
Nevertheless, Face Recognition Vendor Identix®, a company based in Minnesota,
Test 2002 shown that most of these is one of many developers of facial
approaches encountered problems in recognition technology. Its software,
outdoor conditions. This lowered their FaceIt®, can pick someone's face out of a
reliability compared to state of the art crowd, extract the face from the rest of the
biometrics. scene and compare it to a database of
What is Face Recognition? stored images. In order for this software to
work, it has to know how to differentiate
Face recognition technology is the least between a basic face and the rest of the
intrusive and fastest biometric technology. background. Facial recognition software is
It works with the most obvious individual based on the ability to recognize a face and
identifier –the human face. then measure the various features of the
face.
Instead of requiring people to place their
hand on a reader or precisely position their Every face has numerous, distinguishable
eye in front of a scanner, face recognition landmarks, the different peaks and valleys
systems unobtrusively take pictures of that make up facial features. FaceIt defines
 these landmarks as nodal points. Each
human face has approximately 80 nodal
points. Some of these measured by the
software are:

 Distance between the eyes TECHNOLOGY

 Width of the nose
 Depth of the eye sockets Our technology is based on neural
 The shape of the cheekbones computing and combines the advantages of
 The length of the jaw line elastic and neural networks.
Neural computing provides technical
These nodal points are measured creating a information processing methods that are
numerical code, called a face print, similar to the way information is processed
representing the face in the database. in biological systems, such as the human
brain. They share some key strengths, like
robustness fault-resistance and the ability
to learn from examples. Elastic networks
can compare facial landmarks even if
images are not identical, as is practically
always the case in real-world situations.
Neural networks can learn to recognize
similarities through pattern recognition.

Photo .
FaceIt software compares the face print with other
images in the database.

In the past, facial recognition software has

relied on a 2D image to compare or
identify another 2D image from the
database. To be effective and accurate, the
image captured needed to be of a face that
was looking almost directly at the camera,
with little variance of light or facial
expression from the image in the database.
This created quite a problem.

In most instances the images were not

taken in a controlled environment. Even
the smallest changes in light or orientation
could reduce the effectiveness of the
system, so they couldn't be matched to any
face in the database, leading to a high rate
of failure. In the next section, we will look
at ways to correct the problem.
3D FACIAL RECOGNITION degrees, while with 2D, the head must be
turned at least 35 degrees toward the
camera.

Measurement
The system then measures the curves of
the face on a sub-millimeter (or
microwave) scale and creates a template.

A newly-emerging trend in facial

recognition software uses a 3D model,
which claims to provide more accuracy.
Capturing a real-time 3D image of a
person's facial surface, 3D facial
recognition uses distinctive features of the
face -- where rigid tissue and bone is most
apparent, such as the curves of the eye
socket, nose and chin -- to identify the
subject. These areas are all unique and
don't change over time.

Using depth and an axis of measurement

that is not affected by lighting, 3D facial
recognition can even be used in darkness
and has the ability to recognize a subject at
different view angles with the potential to
recognize up to 90 degrees (a face in
profile).
Representation
Using the 3D software, the system goes The system translates the template into a
through a series of steps to verify the unique code. This coding gives each
identity of an individual. template a set of numbers to represent the
features on a subject's face.
Detection
Acquiring an image can be accomplished Matching
by digitally scanning an existing If the image is 3D and the database
photograph (2D) or by using a video image contains 3D images, then matching will
to acquire a live picture of a subject (3D). take place without any changes being
made to the image. However, there is a
Alignment challenge currently facing databases that
are still in 2D images. 3D provides a live,
Once it detects a face, the system moving variable subject being compared to
determines the head's position, size and a flat, stable image. New technology is
pose. As stated earlier, the subject has the addressing this challenge. When a 3D
potential to be recognized up to 90 image is taken, different points are
identified. For example, the outside of the Next, we'll look at how skin biometrics
eye, the inside of the eye and the tip of the can help verify matches.
nose will be pulled out and measured.
Once those measurements are in place, an
algorithm will be applied to the image to SURFACETEXTURE
convert it to a 2D image. After conversion, ANALYSIS
the software will then compare the image The image may not always be verified or
with the 2D images in the database to find identified in facial recognition alone.
a potential match. Identix® has created a new product to help
with precision. The development of
FaceIt®Argus uses skin biometrics, the
uniqueness of skin texture, to yield even
more accurate results.

The process, called Surface Texture

Analysis, works much the same way facial
recognition does. A picture is taken of a
patch of skin, called a skin print. That
patch is then broken up into smaller
blocks. Using algorithms to turn the patch
into a mathematical, measurable space, the
system will then distinguish any lines,
pores and the actual skin texture. It can
identify differences between identical
twins, which is not yet possible using
facial recognition software alone.
According to Indenter, by combining facial
The surface texture analysis (STA) recognition with surface texture analysis,
algorithm operates on the top percentage accurate identification can increase by 20
of results as determined by the local to 25 percent.
feature analysis. STA creates a skin print
and performs either a 1:1 or 1:N match FaceIt currently uses three different
depending on wether you're looking for templates to confirm or identify the
verification or identification. subject: vector, local feature analysis and
surface texture analysis.
Verification or Identification
In verification, an image is matched to  The vector template is very small
only one image in the database (1:1). For and is used for rapid searching over
example, an image taken of a subject may the entire database primarily for
be matched to an image in the Department one-to-many searching.
of Motor Vehicles database to verify the  The local feature analysis (LFA)
subject is who he says he is. If template performs a secondary
identification is the goal, then the image is search of ordered matches
compared to all images in the database following the vector template.
resulting in a score for each potential  The surface texture analysis (STA)
match (1:N). In this instance, you may take is the largest of the three. It
an image and compare it to a database of performs a final pass after the LFA
mug shots to identify who the subject is. template search, relying on the skin
features in the image, which
 contains the most detailed front of the screen. Once the user moves
information. out of the line of sight, the computer is
automatically secured from other users.
By combining all three templates, FaceIt®
has an advantage over other systems. It is
relatively insensitive to changes in CURRENT & FUTURE USES OF
expression, including blinking, frowning FACIAL RECOGNITION
or smiling and has the ability to SYSTEMS
compensate for mustache or beard growth
and the appearance of eyeglasses. The
system is also uniform with respect to race
and gender. The ideal solution
All of this makes face recognition ideal for
high traffic areas open to the general
public, such as:

- Airports and railway stations

- Casinos
- Cash points
- Stadiums
- Public transportation
- Financial institutions
- Government offices
- Businesses of all kinds
However, it is not a perfect system. There In the past, the primary users of facial
are some factors that could get in the way recognition software have been law
of recognition, including: enforcement agencies, who used the
system to capture random faces in crowds.
 Significant glare on eyeglasses or Some government agencies have also been
wearing sunglasses using the systems for security and to
 Long hair obscuring the central eliminate voter fraud. The U.S.
part of the face government has recently begun a program
 Poor lighting that would cause the called US-VISIT (United States Visitor
face to be over- or under-exposed and Immigrant Status Indicator
 Lack of resolution (image was Technology), aimed at foreign travelers
taken too far away) gaining entry to the United States. When a
foreign traveler receives his visa, he will
Identix isn't the only company with facial submit fingerprints and have his
recognition systems available. While most photograph taken. The fingerprints and
work the same way FaceIt does, there are photograph are checked against a database
some variations. For example, a company of known criminals and suspected
called Animetrix, Inc. has a product called terrorists. When the traveler arrives in the
FACEngine ID® SetLight that can correct United States at the port of entry, those
lighting conditions that cannot normally be same fingerprints and photographs will be
used, reducing the risk of false matches. used to verify that the person who received
Sensible Vision, Inc. has a product that can the visa is the same person attempting to
secure a computer using facial recognition. gain entry.
The computer will only power on and stay
accessible as long as the correct user is in
 Hard to fool identification number (PIN) to
verify a customer's identity. This
Face recognition is also very difficult to way business can prevent fraud
fool. It works by comparing facial from occurring.
landmarks - specific proportions and
angles of defined facial features - which DRAWBACKS OF THIS
cannot easily be concealed by beards, TECHNOLOGY
eyeglasses or makeup.
While all the examples above work with
the permission of the individual, not all
However, there are now many more systems are used with your knowledge.
situations where the software is becoming
popular. These systems were taking pictures of all
 As the systems become less visitors without their knowledge or their
expensive, making their use more permission. Opponents of the systems note
widespread. that while they do provide security in some
 They are now compatible with instances, it is not enough to override a
cameras and computers that are sense of liberty and freedom.
already in use by banks and
airports. Registered Traveler Many feel that privacy infringement is too
program will provide speedy great with the use of these systems, but
security screening for passengers their concerns don't end there.
who volunteer information. At the
airport there will be specific lines They also point out the risk involved with
for the Registered Traveler to go identity theft. Even facial recognition
through that will move more corporations admit that the more use the
quickly, verifying the traveler by technology gets, the higher the likelihood
their facial features. of identity theft or fraud.
 Other potential applications include
ATM and check-cashing security. CONCLUSION
After a customer consents, the
ATM or check-cashing kiosk As with many developing technologies,
captures a digital image of him. the incredible potential of facial
The FaceIt software then generates recognition comes with some drawbacks,
a face print of the photograph to but manufacturers are striving to enhance
protect customers against identity the usability and accuracy of the systems.
theft and fraudulent transactions. Face recognition promises latest security
 By using the facial recognition invents in the upcoming trends based on
software, there's no need for a bio-metrics and pattern matching
picture ID, bankcard or personal techniques and algorithms.

REFERENCES:
 "Using your body as a key; legal aspects of biometrics".
 Biometrics Consortuim.
 Y. Adini, Y. Moses, and S. Ullman, “Face recognition: the problem of
compensating for changes in illumination direction,” IEEE Trans. Pattern
Anal. Machine Intell., vol. 19, pp. 721–732, July 1997.