2. Contents
Introduction
Development
to Facial Future
of Facial Applications Survey results Conclusion
Recognition development
Recognition
System
WAT Tsing Hei Joshua ZHAO Fangxin CHUI Loksze
3. Introduction to Facial Recognition System
A Computer Application
Automatic identification or verification or a person
or an object
Can be used on a digital image or a video
Compares selected facial features from the image
and a facial database
4. Developement
Pioneers of Automated Facial Recognition: Woody
Bledsoe, Helen Chan Wolf, and Charles Bisson
Development funded by an unnamed intelligence agency
that did not allow much publicity
Recogition with a “graphics tablet” (GRAFACON or
RAND TABLET)
Development difficulties: great variability in head
rotation and tilt, lighting intensity and angle, facial
expression, aging, distance etc.
overcome by normalizing distances to represent the
face in a frontal orientation through computer
5. Types of Facial Recognition system
3-dimensional
Traditional Skin texture analysis
recognition
• Some facial recognition • A newly emerging • Another emerging
algorithms identify trend, claimed to trend uses the visual
facial features by achieve improved details of the skin, as
extracting accuracies, is three- captured in standard
landmarks, or dimensional face digital or scanned
features, from an recognition . This images. This
image of the subject's technique uses 3D technique, called skin
face. Other algorithms sensors to capture texture analysis, turns
normalize a gallery of information about the the unique lines,
face images and then shape of a face. This patterns, and spots
compress the face information is then apparent in a person's
data, only saving the used to identify skin into a
data in the image that distinctive features on mathematical space.
is useful for face the surface of a
detection. A probe face, such as the
image is then contour of the eye
compared with the sockets, nose, and
face data chin.
6. Software using facial
recognition
Google's Picasa digital image
organizer has a built in face
Sony's Picture Motion Browser (PMB)
analyses photo, associates photos
recognition system . It can associate with identical faces so that they can
faces with persons, so that queries be tagged accordingly, and
can be run on pictures to return all differentiates between photos with
pictures with a specific group of one person, many persons and
people together. nobody.
Facebook includes face recognition Windows Live Photo Gallery includes
technology face recognition
7. Developed applications Undeveloped applications
Used in security Used in other Biometric usages Other usages
systems systems
Applications of facial recognition
Trial a facial Use for visa Used for a portable Modern digital
recognition system processing device to assist cameras often
built into borough- people with incorporate a facial
wide CCTV system prosopagnosia in detection system
recognizing their that allows the
acquaintances camera to focus
Use facial Use by casinos to Use as a security and measure
recognition catch card counters measure at ATM's, exposure on the
software to search and other which would face of the subject,
for potential blacklisted capture an image of thus guaranteeing a
criminals and individuals your face, and focused portrait of
terrorists compare it to your the person being
photo in the bank photographed.
Used facial database to confirm
recognition your identity.
software to prevent
voter fraud
10. Online Survey
Target:
-Hong Kong citizens
Information:
-How common is the facial recognition system
-Where do people want to use this
Site:
https://qtrial.qualtrics.com/SE/?SID=SV_74K8wZP8d
QMIUrX
11. Survey Result
Have you ever use the Facial Recognition System before?
30
25
20
15 No
10 Yes
5
0
1
Facial Recognition is not so common,
Only half of the people use the system before
12. Survey Result
Pretend you have not used this system before, which places do you want to
use it ?
12
10
8
Yes
6
No
4
2
0
Laptop / Computer. School attendance. Camera / Mobile Entrance of a place. Crime detection. Entrance of a place.
phone.
Both two groups mainly want try using the system in laptop/computers.
13. Further Development
-Face Recognition Grand Challenge(FRGC) was created in
2006 and becomes 10 times more accurate than before. It
can recognize the identical twins
-The error rate decreased by one-half every two years
14. Video
http://www.youtube.com/watch?v=RWJ93Mtke7I&feature=
15. Conclusion
Positive:
-Fast and convenient in identifying a person
- Great use in society:
- crime detection
- security use
Negative:
- Errors rate is still not satisfactory
- Privacy - personal biometric data kept by the system
- Change of facial characteristics – wearing of glasses