Scribd
Upload
72 views25 pages

Face Rec

The document describes a face recognition system that uses neural networks. It lists the team members and advisor. It then outlines the overview, existing systems problems, and proposed system advantages. The system will use content-based and color-based methods to extract facial features from images and recognize faces based on a neural network model. It details the architectural design including image extraction, matrix creation, and neural network steps. The system is intended to be used for security applications and its requirements are provided.

Uploaded by

SahilaKhan
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
72 views25 pages

Face Rec

The document describes a face recognition system that uses neural networks. It lists the team members and advisor. It then outlines the overview, existing systems problems, and proposed system advantages. The system will use content-based and color-based methods to extract facial features from images and recognize faces based on a neural network model. It details the architectural design including image extraction, matrix creation, and neural network steps. The system is intended to be used for security applications and its requirements are provided.

Uploaded by

SahilaKhan
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 25

Face Recognition

System

Team members:
Rania samreen(04088097)
Sahila khan (04088130)

Guided By :
Prof. Dr. Eshwar Tenneti
 Overview
 Neural Networks
 Existing System
 Proposed System
 Introduction to Project
 Content based method
 Color based method
 Design goals
 Architectural design
 Back algorithm
 applications
 Requirements
 References
oWhat is Face Recognition?
Identify some unique feature in the face
image of a person
o
How are we going to Recognize?
The system will verify the identity of an
applicant by using neural networks
A neural network is a weight directed graph
that models information processing in the
human brain.
 A neural network usually involves a large
number of processors operating in parallel,
each with its own small sphere of knowledge
and access to data in its local memory.
Problems with existing
system:
o Short Term Memory
o Tough Process
o Time Consuming
o Less Secure
Advantages obtained
through face
recognition system
o Do Not Have to
Memorize
o Easy Process
o Fastest Process
o Secure
 The identification of individuals using face
recognition represents a challenging task
with many applications in everyday life as
well as in high security applications.
 There are two basic face detection
techniques:
 Content-based methods and color-based
methods.
oContent based face recognition system extracts the features of the face like

nose, mouth, lips and calculate the distances between them.


oStatistical model of mutual distance between facial features are used to

locate face in the image.


 By using color based
method we can take
into consideration
color based photos.
 content based method
takes up only gray
scale image.
The design affects the performance, reliability
and maintainability.
1. Performance
2. Reliable
3. Maintainability
The entire system can be mainly subdivided into
three parts.

i) Extraction of eye from the face image(EEF)

ii) Creation of image matrix from the eye(CIM)

iii) Neural network


 The input to the system is the face image of
the applicant.
 Scan the image and store it in a directory.
 using a class we will
grab pixels from the
extracted portion of
the image .
 we are calculating RGB
values for each pixel.
 the matrix will be a
binary value of either 0
or 1.
 The input is an image matrix
 number of values of image

number of input neurons.


 The input layer neurons
receive a finite number of
inputs and then will computes
the weighted sum using the
corresponding weights.
 The output is 0 or 1
First, assumes random weights to the input layer and then
update the weight by calculating the error occurred. The error
is the difference between the desired output and the actual
output. The Learning rule used is error correction learning.
Calculate delta in each stepnand network tends to correct that
error. The weights corresponding to each input is stored in a
file. The training data consists of a number of input data’s. The
weights must be satisfiable for all the trained input data’s. The
input values multiplied with weights, the resultant will be a
binary string that will points to a particular string.
 For a neuron with activation function the
delta rule for 's th weight is given by
 By following the above
steps the face can be
recognized based on the
eyes of the person.

 Face recognition can also be done using other


features like lips or nose but we are opting for
eyes as it gives more accuracy in output.
Face recognition system is mainly used for
security reasons:
Passports
Airports
time attendance system
biometric access control system
visitor management system
Scanning for criminals
 Front End: C# .NET

Hardware Requirements:

 Hard Disk : 40 GB
 Ram : 256 MB
 Main Processor : Pentium IV
 Display Type : High Color 800 by
600
 H. A. Rowley, S. Baluja, and T. Kanade, "Neural Network Based Face
Detection", IEEE Transactions on Pattern Analysis and Machine
Intelligence, 20,January, pp. 23-38, 1998.
 “Eye Identification for Face Recognition with Neural Networks”
http://www.ia.hiof.no/prosjekter/hoit/html/nr2_96/eye_id.html
 http://www.neurosolutions.com/products/ns/whatisNN.html
 http://www.neurotechnology.com/
 http://http://ezinearticles.com/?Biometric-Face-Recognition-System---
Face-Recognition-Now-and-Then&id=5000156
 http://www.face-rec.org

You might also like