A mini project presentation on
Face recognition using principal component
analysis
by
ABHILASH KOTAWAR VENKATA NARAYANA CHETTELA KOMIRISHETTI SRAVAN
In today's networked world, the need to maintain the security of information is becoming both increasingly important and increasingly difficult.
BIOMETRICS represents a good compromise between what’s socially acceptable and what’s reliable, even when operating under controlled conditions.
Recently, technology became available to allow verification of "true" individual identity. This technology is based in a field called "biometrics".
INTRODUCTION
TYPES OF BIOMETRICS1.SIGNATURE VERIFICATION2.SPEAKER RECOGNITION3.FINGERPRINT RECOGNITION4.IRIS RECOGNITION5.FACE RECOGNITION BLOCK DIAGRAM FOR RECOGNTION
Face Recognition is the process of identification of a person by their facial image. This technique makes it possible to use the facial images of a person to authenticate him into a secure system, for criminal identification, for passport verification,...
Face recognition technology is the least intrusive and fastest biometric technology.
Face recognition systems unobtrusively take pictures of people's faces as they enter a defined area.
This method is found to be fast, relatively simple, and works well in a constrained environment.
WHY FACE RECOGNITION?
HOW DOES FACE RECOGNITION WORKS?
PCA is a dimensionality reduction technique based on extracting the desired number of principal components of the multi-dimensional data.
PCA aims to: Summerise data with many independent
variables to a smaller set of derived variables.
identifying patterns in data, and expressing the data in such a way as to highlight
their similarities and differences.
PRINCIPAL COMPONENT ANALYSIS
Get some data:
Mean=∑ Xi/n
variance=(∑(xi-avg)²)*1/(n-1) sum of variances=16.3756
THE PROCESS OF PCA
x y
1.4000 1.6500
1.6000 1.9750
-1.4000 -1.7750
-2.0000 -2.5250
-3.0000 -3.9500
2.4000 3.0750
1.5000 2.0250
2.3000 2.7500
-3.2000 -4.0500
-4.1000 -4.8500
Average -0.4500 -0.5675
Variance 6.4228 9.9528
For covariance we will use function (∑(x-xbar)*(y-ybar)/(n-1)
Calculating covariance matrix
X-Xbar Y-Ybar (X-Xbar)*(Y-Ybar)
1.8500 2.2175 4.1024
2.0500 2.5425 5.2121
-0.9500 -1.2075 1.1471
-1.5500 -1.9575 3.0341
-2.5500 -3.3825 8.6254
2.8500 3.6425 10.3811
1.9500 2.5925 5.0554
2.7500 3.3175 9.1231
-2.7500 -3.4825 9.5769
-3.6500 -3.4825 15.6311
7.9876 covariance
In general the covariance matrix is = [covariance(x,x) covariance(x,y) covariance(y,x) covariance(y,y)] = [variance(x) covariance(x,y) covariance(x,y) variance(y)] = [6.4228 7.9876 7.9876 9.9528]
To obtain Eigen values by solving function determinant {A-lamda(I)}=0
Solving equation A, we get the Eigen values are lamda=16.36809984,0.007462657
Here sum of two eigen values is always equal to the sum of variances
To obtain Eigen vector by solving for matrix x in such a way that, {A-lambda(i)}*[X]=[0].
For first Eigen value 16.36809984, we get [X]=[0.6262
0.7797] For second Eigen value 0.007462657,we get [X]=[0.7797 -0.6262]To obtain coordinates of data point in the
direction of Eigen vectors by multiplying the centered data matrix to the Eigen vector matrix
X-Xbar Y-Ybar
1.8500 2.2175
2.0500 2.5425
-0.9500 -1.2075
-1.5500 -1.9575
-2.5500 -3.3825
2.8500 3.6325
1.9500 2.5925
2.7500 3.3175
-2.7500 -3.4825
-3.6500 -4.2825
= [0.62620 0.77967 0.77970 0.62620] =
Projection on the line of first principal component
Projection on the line of second principal component
2.88737 0.505380
3.26600 0.00622
-1.53633 0.01545
-2.49680 0.01729
-4.23402 0.12995
4.62439 0.05886
3.24237 0.10306
4.30858 0.06669
-4.43722 0.03664
-5.62453 0.16411
16.36809775 0.007462657
STEP1.Get some dataSTEP2.subtract the meanSTEP3.Calculate the covariance matrixSTEP4.Calculate the Eigen vectors & Eigen values
of the covariance matrixSTEP5. choosing components and forming a
feature vector The variance of projections in the line of principal
component is equal to the Eigen values of the principal components.
First Eigen vector is able to explain around 99% of total variance
SUMMARY OF PCA
DATABASE PREPATATION TRAINING TESTING
Flow chart indicating the sequence of implementation
IMPLEMENTATION OF FACE RECOGNITION USING PCA IN MATLAB
Flow chart for training & testing
1.Acess control ATM AIRPORT A door lock control system
2.Entertainment: Video Game Human Computer Interaction Human Robotics
APPLICATION AREAS
3 Smart cards: Driver’s license Passports Voter registrations Pan card 4 Information Security: Desktop Logon Personal Driven Logon Database security 5 law Enforcement And Surveillance: Advanced video surveillance Drug trafficking
And some other Commercial Applications:
APPLICATION AREAS
Face recognition is also very difficult to fool. It works by comparing facial and marks - specific proportions and angles of defined facial features - which cannot easily be concealed by beards, makeup.
By using the facial recognition software, there's no need for a picture ID, bankcard or personal identification number (PIN) to verify a customer's identity. This way business can prevent fraud from occurring.
HARD TO FOOL
A face needs to be well lighted by controlled light sources in automated face authentication systems. This is only a first challenge in a long list of technical challenges that are associated with robust face authentication.
The risk involved with identity theft.
DRAWBACKS
Face recognition is a both challenging and important recognition technique. Among all the biometric techniques, face recognition approach possesses one great advantage, which is its user-friendliness.
Face recognition promises latest security invents in the upcoming trends based on bio-metrics and pattern matching techniques and algorithms.
CONCLUSION
CONCLUSION:
The image may not always be identified in facial recognition alone.
A picture is taken of a patch of skin, & is then broken up into smaller blocks, Using algorithms.
It can identify differences between
identical twins, which is not yet possible
using facial recognition software.
Accurate identification can increase by 20 to 25 percent.
SURFACE TEXTURE ANALYSIS