E017443136

IOSR Journal of Computer Engineering (IOSR-JCE)
e-ISSN: 2278-0661,p-ISSN: 2278-8727, Volume 17, Issue 4, Ver. IV (July – Aug. 2015), PP 31-36
www.iosrjournals.org
DOI: 10.9790/0661-17443136 www.iosrjournals.org 31 | Page
Analysis of Iris Recognition Based On FAR and FRR
Using Hough Transform
Simranjit Kaur, Sourav Garg
(CSE, ACET/PTU,Jalandhar, India)
(AP in CSE, ACET/PTU,Jalandhar, India)
Abstract: Iris recognition is an identification method of biometric that uses pattern-recognition techniques. It
is one of the most biometrical techniques used for personal identification. In this paper, we give a brief overview
of different methods used in iris recognition system and use houghman transform and rubber sheet method for
increasing accuracy by calculating metrics like FAR and FRR.
Keywords: Biometrics, Canny edge detection, FAR, FRR, Houghman Transform, Normalization Basic Rubber
Sheet Model
I. Introduction
Biometrics is a system that measure physical and behavioral uniqueness to identify an individual.
Instead of requiring personal identification cards, keys or passwords, biometrics can identify fingerprints, face,
iris, signature, or retinas of an individual for easy verification. Iris Recognition is a biometrical technology for
personal identification and verification which recognize a person from his/her iris prints.
Figure.1 Eye
II. Stages In Iris Recognition System
A. Image Acquisition- It is for capturing a number of iris images using a designed sensor. Image acquisition
step is one of the most sensitive and important for the quality of image to be processed, data extracted from raw
input determines the performance of the entire system to a large extent.
B. Localization/Segmentation- The iris is acquired as a part of a larger image that contains data derived from
the surrounding eye region. The inner and the outer boundaries of the iris are calculated.So it is important to
localize that portion of the image that corresponds to iris.
C. Normalization- It is that process in which there is change in the range of pixel intensity values. It produce
iris region which have the same constant dimensions, so that two images of the same iris under different
conditions will have same features.
D. Feature Extraction- The most important step in iris recognition is the ability of extracting some unique
attributes from iris, which help to generate a specific code for each individual. Feature extraction is a special
form of dimensionality reduction.
E. Pattern Matching-Pattern Matching phase is the last phase after all phases. The feature which is generated
in the feature extraction method is used as the input value for comparing the iris image with the iris codes.
Figure.2 Stages
LOCALIZATION NORMALIZATION
PATTERN MATCHING
FEATURE EXTRACTION
ACCEPT
REJECT
IMAGE ACQUISITION

Analysis Of Iris Recognition Based On Far And Frr Using Hough Transform
III. Methods Used In Existing Iris Recognition
A. Localization or Segmentation
a. Canny Edge Detection: It is an edge detection operator that is used to detect a wide range of edges in
images. The boundary coordinates are separated from the rest of the image based on the intensity value of the
pixel.
b. Hough Transform: It is used for identify lines in the image. It is used to estimate the parameters of a shape
from its boundary points in a image.
c. Circular Hough Transform: It is used to deduce the radius and centre coordinates of the pupil and iris
regions by biasing the derivatives in horizontal direction during edge mapping detects the eyelids and biasing
the derivatives in the vertical direction helps in detecting the circular boundary of the iris [7].
d. Daughman’s Intergo-Differertial Operator: It is used for locating the circular iris and pupil region. This
operator searches for the circular path. It takes the close up image as input from that where there is maximum
change in pixel values by varying the radius and centre(x & y) positions.
B. Normalization
a. Daughman’s Rubber Sheet Model: It converts the Cartesian coordinates (e.g. [1, 2]) into polar coordinates
(e.g. [radius, angle]).It takes into account pupil dilation and size consistencies to produce normalized
representation with constant dimensions [1].
b. Rubber sheet model of lower half of iris: The iris is partially covered with eyelids. This can lead to false
recognition result. To avoid the prominent occlusion due to upper eyelid this technique was
Implemented [1].
c. Sector Based Normalization: There exists traditional method for iris normalization that transforms the
detected annular ring into a rectangular block. The sector based normalization helps to minimize by eyelashes
and eyelids. For certain range of angular values the radial dimension along left and right sector are taken
completely because no occlusion occurs here. However for the upper and lower region only partial values are
taken in the sector.
d. Analysis Band Based Normalization: In this technique of normalization the portion of the iris to be
analyzed is mapped and subdivided into eight analysis bands [1].
e. Dimension Reduction Approach: This approach is used to reduce the dimensionality of the problem. It takes
the two portions having the size equal to the pupil size from the left and right side of the iris. After performing
normalization operation, these two portions are merged [1].
C. Feature Extraction
a. Cumsum Based CPA: This is a algorithm used for extract features by generating iris codes by analyzing
changes of grey values of iris pattern by using change point analysis which determine whether a change has take
place from darkness to brightness or brightness to darkness.
b. Daubechies Wavelets: It is used to transform huge data in smaller representations by using filters that
produce smooth and scaling functions.
c. Ridgelet: It is combination of random transform and wavelet transform. It relates the scales of the line
position with wavelet transform and relates it to the scale of the point positions.
D. Pattern Matching
a. Hamming Distance: It is a method that calculates the bit difference by using the XOR operation on the iris
patterns. Hamming distance between two strings of equal length is the number of positions at which
corresponding symbols are different.
b. Euclidean Distance: Euclidean distance is the matching metric to match the iris with the iris images already
present in the data base.

c. K-NN Classifier: It is one of the most important non-parameter algorithms and it is a supervised learning
algorithm. The classification rules are generated by the training samples themselves without any additional data.
The KNN classification algorithm predicts the test sample’s category according to the K training samples which
are the nearest neighbors to the test sample, and judge it to that category which has the largest category
probability [5].
d. Histogram Matching: It is a graphical representation of the tonal distribution in a digital image. It plots the
number of pixels for each tonal value. Pattern matching is performed by this technique.
E. Accept/Reject
Final phase is for accepting the code or rejecting the code. It depends on identification and verification which
are important for security of every system. In the verification, the system checks if the user data that was entered
is correct or not (e.g., username and password) but in the identification stage, the system tries to find who is
without any input information.
IV. Proposed Methodology
In this proposed system we use following steps with Hough transform and rubber sheet techniques.
A. Input Data:-In first step there is capturing of image from cassia database.
B. Define Length:-In second step length is calculated.
C. Edge Detector (canny edge detector):- It is an edge detection operator that uses a multistage algorithm to
detect a wide range of edges in images. It is a method of segmentation. The boundary coordinates are separated
from the rest of the image based on the intensity value of the pixel and strength of them.
D. Boundary Detector(Hough transform):- It is used for identify lines in the image. It is used to estimate the
parameters of a shape from its boundary points in a image. In this there are three types of detection:-1.Pupil
region detection.2.Pupil boundary.3.Outer boundary detection.
E. Find Boundary Coordinates:-In this step there is searching of boundary coordinates.
F. Detection of noisy region and noisy circles:-In next steps there is detection of noisy parts like noisy regions
and noisy circles.
G. Normalization (Rubber sheet):- It converts the Cartesian coordinates ( [1, 2]) into polar coordinates (
[radius, angle]).It takes into account pupil dilation and size consistencies to produce normalized representation
with constant dimensions [1].
H. Histogram:- Histogram that acts as a graphical representation of the tonal distribution in a digital image. It
plots the number of pixels for each tonal value. The histogram for a specific image a viewer will be able to
judge the entire tonal distribution at a glance. Image histograms are present on many modern digital cameras. It
evaluates the accuracy of images.
I. Output values:-There are some values used to determine performance.
a. Mean Squared Error MSE is a risk function, corresponding to the expected value of the squared error loss
or quadratic loss. MSE measures the average of the squares of the "errors."
b. Root Mean Squared Error:- RMSE is used to measure differences between values predicted by a model or
an estimator and the values actually observed. It is the square root of mean squared error.
c. Signal-to-noise ratio:-SNR is a measure that compares the level of desired signal to the level of background
noise. It also refers to the ratio of useful information to false data in a exchange.
d. Root Signal-to-noise ratio:- It is a square root of measurement that compares the level of desired signal to
the level of background noise. It also refers to the ratio of useful information to false data in a exchange.
e. Peak signal-to-noise ratio:-PSNR is an engineering term for the ratio between the maximum possible power
of a signal and the power of corrupting noise that affects the fidelity of its representation.

Figure.3Methodology
V. Performance Evaluation
Following metrics are used to evaluate the performance of the system.
1. False Acceptance Rate (FAR): FAR is the measure of the likelihood that the biometric security system will
incorrectly accept an access attempt by an unauthorized user. FAR is defined as the ratio of the number of
false acceptances divided by the total number of identification attempts.Accuracy increases with decrease in
FAR.
2. False Rejection Rate (FRR): FRR is the measure of the likelihood that the biometric security system will
incorrectly reject an access attempt by an authorized user. FRR is defined as the ratio of the number of false
rejections divided by the total number of identification attempts.Accuracy increases with decrease in FRR.
VI. Result
The proposed technique has been applied to the following database(CASIA):
Images:-
Figure.4 Database Images

Far And Frr:-
Figure.5 Result of FAR and FRR
Mse,Rmse increases—accuracy decreases,Mse, Rmse decreases—accuracy increases
Psnr,Snr increases—accuracy increases ,Psnr,Snr decreases—accuracy decreases
FAR increases—accuracy decreases, FAR decreases—accuracy increases
FRR increases—accuracy decreases, FRR decreases—accuracy increases
Comparision With Existing Systems:-
System[reference] FAR FRR Overall % Accuracy
Avila 0.03 2.08 82.2%
Li Ma 0.02 1.98 83.4%
Tisse 1.84 8.79 74.5%
Daugman 0.01 0.09 88.1%
Using Circular Hough Transform 0.005 0.238 95.6%
Proposed System
CASIA
0.013 0.018 99.3%
Figure.6 Accuracy
VII. Conclusion And Future Scope
We have study all methods and approaches on some images which are used in iris recognition system
and compare them according to their values of FAR AND FRR. The images which have low values of FAR and
FRR, they have more accuracy then other images. We have study all methods which are used in iris recognition
system and compare them according to their performance. Techniques and methods can be improved which are
used under system for increasing the accuracy of iris recognition system. This can be done by decreasing FAR
and FRR, so that final outcome can be improved.

References
[1]. Namrata P. Joshi, Roopal K. Lamba, Devang U. Shah, Bhargav V. Ghadia, Implementation of Various Approaches for Iris Image
Normalization, Institute Of Technology, Nirma University, Ahmedabad –382 481,IEEE,08-10 dec, 2011
[2]. J. Daugman, „How iris recognition works‟, Proceedings of 2002 International Conference on Image Processing, Vol. 1, 2002.
[3]. Emmanvel Raj.M.Chirchi, Dr.R.D.Kharadkar, Improvement of Iris Pattern Authentication System using Cumulative Sum
Algorithm, International Journal of Engineering and Innovative Technology, ISSN: 2277-3754, Vol 2, Issue 5, Nov 2012.
[4]. Anu Sharma Dr. Vijay Kumar Banga, Improving Performance Of An Iris Recognition System Using Correlation Based Matching,
International Journal of Engineering Research & Technology (IJERT) ISSN: 2278-0181 Vol. 2 Issue 4, April -2013.
[5]. Dolly Choudhary, Ajay Kumar Singh, Shamik Tiwari, A Statistical Approach for Iris Recognition Using K-NN Classifier, I.J.
Image, Graphics and Signal Processing, 2013, 4, 46-52 Published Online April 2013 in MECS.
[6]. Alinka P. Takale, Vanaja Roselin. E.Chirchi, Iris Recognition: Cumsum Based CPA Approach, -International Journal of Advanced
Computational Engineering and Networking, ISSN: 2320-2106, Volume- 1, Issue- 5, July-2013.
[7]. Matsoso Samuel Monaheng, Padmaja Kuruba, Iris Recognition Using Circular Hough Transform, International Journal of
Innovative Research in Science, Engineering and Technology, ISSN: 2319-8753, Vol. 2, Issue 8, Aug 2013.
[8]. Bansal,A; G.L.A.Univ,Mathura,India;Agarwal,R.;Sharma,R.K, FAR and FRR based analysis of iris recognition system, IEEE, 15-
17 March 2012.
[9]. Simranjit Kaur, Jasmeen, ―Survey of Different Approaches in Biometric Iris Recognition System‖, International Journal of
Advance Research in Computer Science and Software Engineering, Vol-4, Issue 7 July 2014, ISSN: 2277 128X 768-7710.
[10]. Simranjit Kaur, Garg Sourav, ―Analysis of Iris Recognition System by using different methods‖, National Conference on Emerging
Trends in Engineering & Technology 7-8 November 2014,ACET Bhawanigarh.
[11]. http://www.icdri.org/biometrics/iris_biometrics.htm
[12]. Center for biometrics and security research. http://www.cbsr.ia.ac.cn/
Books:
[13]. Gonzales Rafeil C., Woods ― Digital Image Processing‖ Edition third, 416-440, 564-626.

E017443136

More Related Content

What's hot (19)

Viewers also liked (20)

Similar to E017443136 (20)

More from IOSR Journals (20)

Recently uploaded (20)

E017443136