Smiling Faces 04/21/2010 2010_Spring_CS4893_Smiling Faces_ ...

herbunalaskaData Management

Jan 30, 2013 (4 years and 8 months ago)

129 views

Smiling
Faces



0
4/21
/2010


Page
1

of
1

Filename: 2010_Spring_CS4893_Smiling Faces_

Project Abstract

_

Ver 1.00


2010_Spring_CS4893_Smiling Faces_

Project Abstract

Version:
1.00

Facial Recognition is the process of identifying a person by matching a captured image of a face to a
record image

of that person. O
ur goal was

to develop a

software application that will
provide
the
following features:



Image capturing

of a face from live video



Storing, recording

of
captured
images in a

open source

database



Matching
captured image

against the images stored in the database

The name of our team

was Smiling Faces

and our team

number was 4.
The members in our team were
Nirav Sheth, Evan Jacobs, Jason Brand, Anusha Gattupalli, and Benjamin Duggar.

The application was coded in C++, using Microsoft Foundation Class graphical user interface (GUI)
features in Visual Studio 2008. Ima
ge capturing from live video was done using Microsoft DirectShow
library. Face detection
and recognition
was done
using an open source computer vision library called
OpenCV
. The image record database
was

created using an open source database management sys
tem
(DMBS) PostgreSQL
; and the database access library used was libpq 8.4
.
Incremental
s
oftware

l
ife
c
ycle
model

was
used to develop this application
, because our team started out with limited knowledge of all
the components of this application. Thus indiv
idual components were created and tested in isolation
and then integrated together in an incremental way.

The resulting application is a robust face recognition system that can be used for small to medium scale
face recognition tasks in a controlled envir
onment
;

such as gateway authentication,
self
-
service kiosk
security, etc.
In conclusion, we have created a working face recognition application that can be easily
deployed
, operated, and
used to

recognize
faces.