Abstract
Most of the biometric authentication systems store Past Work Done The recognition ability of classifiers
multiple templates per user to account for variations depends on the quality of feature used as well as the
in biometric data. Therefore, these systems suffer amount of training data available to them. Image
from storage space and computation overheads. To features are mostly extracted on shape and texture of
overcome this problem the paper proposes techniques segmented objects. Karu et al. [15] define a
to automatically select prototype templates from iris methodology for automatically identifying textured
textures. The paper has two phases: one is to find the regions within an image so that feature extraction
feature vectors from iris textures that have less algorithms are only used where texture can be
correlation and the second to calculate DU measure. quantified. The classification is mainly based on
DU measure is an effective measure of the similarity statistical measures used for texture classification
between two iris textures, because it takes into such as DU measures, autocorrelation, co-occurrence
consideration three important perspectives: a) matrices etc. Textured methods used can be
information, b) angle and e) energy. Also, gray level categorized as: statistical, geometrical, structural,
co occurrence matrix is used to find the homogeneity model-based and signal processing features
and correlation between the textures. [13].Randen and Husoy conclude that most studies
deal with statistical, model-based and signal
Keywords Shaker iris, Jewel iris, Flower iris, Stream processing techniques. Co-occurrence method is
iris, gray level co-occurrence matrix, Spectral proved best for texture classification by Conners and
information divergence, Spectral angle mapper, DU Harlow [14].
measure
No work has been carried out in classification of iris
1. Introduction textures to overcome the overheads due to space and
time for high speed recognition, until today. This
A typical iris biometric system operates in two paper tries to classify iris textures on guidelines of
distinct stages: the enrollment stage and the statistical texture classification methods as discussed
authentication stage [2]. During enrollment stage iris above.
textures are acquired and processed to extract a The remainder of this paper is organized as follows:
feature set. The complex iris textures carry very Section 2 briefly summarizes basic types of iris
distinctive information. The feature set includes some textures. Section 3 describes the necessity of template
distinguished features like nodes and end point in the selection and updating. Section 4 describes statistical
iris texture. The stored feature set, labeled with the measures used to define the relation between pixels.
user’s identity, is referred as a template. In order to A detailed description of the proposed measurements
account for variations in the biometric based used to classify the textures is given in Section 5.
authentication system relies on the stability from an Section 6 and 7 reports the experiments and discuss
individual is susceptible to changes due to distance the results and Section 8 presents the detailed flow
from the sensor, poor resolution of the sensor and chart of the proposed algorithm. Finally Section 9
alterations in iris texture due to cataract operations presents the conclusion.
and others.
Multiple iris features pertaining to different portions
2. Types of Iris Textures
of iris must be stored in data base. There is a trade off
Depending on the texture of different human iris, we
between the number of templates, and the storage and
can group the irises into four basic groups. They are
computational overheads introduced by multiple
a) Stream iris, b) Jewel iris, c) Shaker iris and d)
templates. An efficient system must in fact classify
Flower iris. Also, there are different combinations of
the templates automatically.
17
ICGST-BIME Journal, Volume 7, Issue 1, May, 2007
there groups. As a preliminary study we have otherwise Flower iris is sufficient to cause the
concentrated to group the given data base into the occupant to exhibit shaker characteristics.
above four classes.
a) Stream Iris: It contains a uniform fiber structure
with subtle variations or streaks of color as shown in
figure 1.c. The structure of the iris is determined by
the arrangement of the white fibers radiating from the
center of the iris ( or pupil). In this image one can
notice that they are uniform and reasonably direct or
parallel.
Figure 3.a & b: Shaker iris
18
ICGST-BIME Journal, Volume 7, Issue 1, May, 2007
Some of the minutiae points like nodes and end d) Homogenity measures the closeness of the
points of iris textures may be added or deleted from distribution of elements in the GLCM to the
the iris template. Thus template selection refers to the GLCM diagonal.
process by which iris textures are chosen from a
given set of samples, where as template update refers We have derived only correlation to identify the
to the process by which existing templates are length of the feature vector which has unique
replaced. information from iris texture. A typical iris exhibits
rich texture information in the immediate vicinity of
4. Selection of feature vectors the pupil which tapers away in intensity as one moves
GLCM(gray-level co-occurrence matrix)[4], also away from the pupil. Thus, the iris templates are
known as the gray-level spatial dependence matrix, is taken by segmenting a portion of iris texture near the
a statistical measure used to characterize the texture pupil and from them the feature vectors are derived,
of an image by calculating how often pairs of pixel, by using correlation function. This is implemented by
with specific values and in a specified spatial performing canny edge detection technique to
relationship occur in an image and then by extracting produce binary image or respective iris texture and
statistical measures from this matrix. also by calculating the probability of occurrence of
By default, the spatial relationship is defined as the pixels.
pixel of interest and the pixel to its immediate right
(horizontally adjacent), but one can specify other
spatial relationships between the two pixels. Each
element (i,j) in the resultant GLCM is simply the sum
of the number of times that the pixel with value I
occurred in the specified spatial relationship to a pixel Figure 5.a Figure 5.b
with value j in the input image.
Initially, the dynamic range of the given image is
scaled to reduce the number of intensity values in
intensity image from 256 to 8. The number of gray-
levels, determine the size of the GLCM.
The gray-level co-occurrence matrix can reveal Figure 5.c Figure 5.d
certain properties about the spatial distribution of the Fig:5 Iris textures after canny edge detection
gray levels in the texture image. For example, if most
of the entries in the GLCM are concentrated along the a) Stream iris b)Jewel iris c0Shaker iris d)Flower
diagonal, the texture is coarse with respect to the iris The correlation Vs offset plot for different
specified offset. iris textures are given below.
However, a single GLCM might not be enough to
describe the textual features of the input image. For
example, a single horizontal offset might not be
sensitive to texture with a vertical orientation. For
this reason, gray co-matrix [5] can create multiple
GLCMs for a single input image. TO create multiple
GLCMs, specify an array of offsets to the gray co-
matrix function. These offsets define pixel
relationships of varying direction and distance. For,
example, one can define an array of offsets that
specify four directions (horizontal, vertical and two
diagonals) and four distances. In this case, the input
image is represented by 16 GLCMs. The statistics are Figure 6.a Correlation of Shaker iris texture
calculated from these GLCMs and then the average is
taken.
19
ICGST-BIME Journal, Volume 7, Issue 1, May, 2007
Figure 6.c Correlation of Stream iris texture Finally DU measure is given by,
6. Iris Classification
When a new iris texture is to be grouped into a
predefined class, the Du measure is calculated with
any one of the typical templates in that class. If the
Du score is very high, we can say that it does not
belong to that class. The new template belongs to a
class, with which its Du measure is very low in
comparison to others [1].
20
ICGST-BIME Journal, Volume 7, Issue 1, May, 2007
SID SAM
Find the relative entropy of each
template,
e1= f*log2(f/g);
e2= g*log2(g/f);
MM=SID*Tan(SAM)
SID=e1+e2
Stop
Stop
d)Flow chart to find APD
If f & g are two iris templates
b) Flow chart to find SAM
Read two iris templates Read two iris templates
to be compared, f &g to be compared, f &g
N size of template N size of template
i=1,j=1
h=f-g;
h1=Σh;
i=N&
j!=N APD=h/N
Stop
f(i,j)=f(i,j)*f(i,j)
g(i,j)=g(i,j)*g(i,j)
fg(i,j)=f(i,j)*g(i,j) 9. Conclusion
In this paper we mainly concentrated on basic
classification of iris textures into four groups. Our
results show that there is a considerable variation in
DU measure for different combinations which are yet
to be standardized with a large data base. Future work
includes further classification of iris textures into sub
21
ICGST-BIME Journal, Volume 7, Issue 1, May, 2007
groups. Also not only DU measures, but statistical [14] R.W.Conners and C.A.Harlow, “A theoretical
texture methods such as auto correlation, edge comparison of texture algorithms”, IEEE
frequency, primitive lengths, Law’s texture features transactions on Pattern Analysis and Machine
can be used for more efficient and accurate Intelligence, vol2,no. 3,pp.204-222, 1980.
classification. [15] K.Karu, A.K.Jain and R.M.Bolle, “Is there any
10. References texture in the image?”, Pattern Recognition, vol.
[1] NIST, Advanced Encryption Standard 29,no. 9,pp-1437-1446,1996.
(AES),2001.http://csrc.nist.gov/publications/fips/
fips-197.pdf Biographies
[2] H.Heijmans, Morphological Image Operators,
AcademiPress, 1994. Srinivas R.E. received the
[3] A.Juels and M.Sudan, “A Fuzzy Vault Scheme”, B.Tech degree in Electronics &
proc. IEEEInt’l. Symp.Inf. Theory,A Communication ENgienering
Lapidothand E.Teletar,Eds.,pp.408,2002. from Nagarjuna University,
[4] R.C.Gonzalez amd R.E.Woods, Digital Image India in 1988, M.S. degree from
Processing, Addison-Wesley, 3rd ed.,1992. Birla Institute of Technology
[5] Joaquim De Mira Jr, Joceli Mayer, “Image and Scince, India in 1997,
Feature Extraction for application of Biometric M.Tech degree in Computer
Identification of Iris – A Morphological Science from Visveswaraiah Technological
Approach”. proc IEEE Int’l Symp on Computer University, India in 2000 and pursuing his Ph.D in
Graphics and Image processing, SIBGRAPI’03 computer science in Nagarjuna Univeristy under the
[6] Umut Uludag, and Anil K.Jain, “ Fuzzy Finger guidance of Dr. I. Ramesh Babu. He is the member of
Print Vault”, Proc. Workshop: Biometrics: IEEE and presented 8 papers in international
Challenges Arising from Theory to practice, conferences. His research interests includes image
pp.13-16,2004,W.H.Press processing, biometrics and pattern recognition
[7] S.A.Teukolsky, W.T.Vetterling, and
B.P.Flannery, Numerical Recipes in C, 2.
Ed., Cambridge University press, 1992. Ramesh Babu.I received the
[8] Chuan Chin Teo, Hong Tat Ewe “An efficient B.Tech degree in Electronics &
One Dimensional Fractal Analysis for Iris Communication Engineering
Recognition”WSCG’2005, January 31 from Mysore University, India in
Feb4,2005,Plzen,Czech Republic. 1981, M.Tech degree in
[9] LiMA,Tieniu, “Efficient Iris Recognition by Computer Engineering from
Characterizing Key Local Variations”, IEEE Andhra University, India in 1984
trans., Image Processing-2004. and Ph.D degree in computer Science & Engineering
[10] Robert Ives, Delores Etter, Yingzi Du, “Iris from Nagarjuna University, India in 1994.
Pattern Extraction using Bit Planes and Standard He is currently working as Head & Professor in the
Deviations”, IEEE conference on Signals, department of computer science, Nagarjuna
systems and computers, 2004. University. Also he is the senate member of the same
[11] Qi-Chuan Tian, Quan Pan, Yong-Mei Cheng“ University from 2006. His areas of interest are image
Fast algorithm and application of Hough processing & its applications, and he is currently
Transform in iris segmentation”, proceedings of supervising 10 ph.D students who are working in
third IEEE conference on machine learning and different areas of image processing. He is the senior
Cybernetics, Shangai, 26-29, 2004 . member of IEEE, and published 35 papers in
[12] Mona Sharma, Markos Markou, Sameer Singh international conferences and journals.
“Evaluaion of Texture Methods For Image
Analaysis”, pattern recognition letters.
[13] T.R.Reed and J.M.H.Buf, “A review of recent
texture segmentation and feature extraction
techniques”,Computer visison, Image Processing
and Graphics, vol.57,no. 3,pp. 359-372,1993.
22