Sie sind auf Seite 1von 6

Novel Approach in Iris Recognition

B. Kiran Bala and A. Josephine Pushpa Arasi M.A.M. College of Engineering Department of Computer and Communication Engineering Trichy, Tamil Nadu, India. Abstract--Iris recognition is an intrinsic human organ, capturing of eye image then locate iris from the eye color image extract the features finally match with database. for theses process the first step to preprocessing the color image, the second step to locate the iris using the canny edge filter which locate the inner boundary of iris and outer boundary of iris can be identify by huge variation in the intensity value, the third step to convert Cartesian form into polar form using rubber sheet modal suggested by Daugman now the iris color image ready to match with database this system having feature extraction like color, texture and edge should be matched with database. The process is for contact lens and disease affected problem also addition to this we are having the iris sharpness and image enhancement technique to overcome the problems. Another important aspect of this system for the color image processing own database has been created for this process under the guideless of doctor and image samples taken from the hospital those who are willing to give there eyes for research. Efficiency of this system can be justified by the result of comparison between gray and color image processing.



The eye is not properly a sphere; rather it is a fused two-piece unit. The smaller frontal unit, more curved, called the cornea is linked to the larger unit called the sclera. The corneal segment is typically about 8 mm (0.3 in) in radius. The sclera constitutes the remaining five-sixths; its radius is typically about 12 mm. The cornea and sclera are connected by a ring called the limbus. The iris the color of the eye and its black center, the pupil, are seen instead of the cornea due to the cornea's transparency. To see inside the eye, an ophthalmoscope is needed, since light is not reflected out. The fundus (area opposite the pupil) shows the characteristic pale optic disk (papilla), where vessels entering the eye pass across and optic nerve fibers depart the globe [1].The dimensions differ among adults by only one or two millimeters. The vertical measure, generally less than the horizontal distance, is about 24 mm among adults, at birth about 1617 mm. (about 0.65 inch)

The eyeball grows rapidly, increasing to 22.523 mm (approx. 0.89 in) by the age of three years. From then to age 13, the eye attains its full size. The volume is 6.5 ml (0.4 cu. in.) and the weight is 7.5 g. (0.25 oz.).The eye is made up of three coats, enclosing three transparent structures. The outermost layer is composed of the cornea and sclera. The middle layer consists of the choroid, ciliary body, and iris. The innermost is the retina, which gets its circulation from the vessels of the choroid as well as the retinal vessels, which can be seen in an ophthalmoscope. Within these coats are the aqueous humor, the vitreous body, and the flexible lens. The aqueous humor is a clear fluid that is contained in two areas: the anterior chamber between the cornea and the iris and exposed area of the lens. The lens is suspended to the ciliary body by the suspensory ligament (Zonule of Zinn), made up of fine transparent fibers. The vitreous body, the posterior chamber, is a clear jelly that is much larger than the aqueous humor, present behind lens and the rest, and is bordered by the sclera, zonule, and lens. They are connected via the pupil [2]. The biometric person authentication technique based on the pattern of the human iris is well suited to be applied to any access control system requiring a high level of security. This paper examines a new iris recognition system that implements (i) gradient decomposed Hough transform / integro-differential operator combination for iris localization and (ii) the analytic image concept (2D Hilbert transform) to extract pertinent information from iris texture. All these imageprocessing algorithms have been validated on noised real iris images database. The proposed innovative technique is computationally effective as well as reliable in terms of recognition rates. Our basic experimentation of the Daugmans Mathematical algorithms for iris processing, derived from the information found in the open literature, led us to suggest a few possible improvements. For justification of these new concepts we implemented in C language both the Daugmans iris processing unit and our own. Afterwards we tested individually the performances of the different processing blocks previously identified as follows: (1) locating iris in the image, (2) Cartesian to polar reference transform,

(3) local features extraction, and (4) matching. Advantage of the natural circular shape to improve the success rate of iris localization. The iris codes of size 768 bytes that we compared in this test included a sampling of the binary images resulting of the emergent frequency functions [8]. 2. PROPOSED SYSTEM

Fig. 1. Basic structure of proposed system

The following process is for the gray scale image processing as well as algorithm to process and various step involved in the gray scale image iris recognition process

A. Locate the Iris in gray scale

First step in this process convert the color image into gray scale then convert into standard size format 256 X 256 gray scale image and then segmentation process is very important to locate the inner and outer boundary of iris from the eye image and then normalization will convert the Cartesian form into polar form. B. Feature extraction in gray scale image After located the iris next step to extract the feature like texture, edge, shape from the image for that we can use several algorithm to locate the texture and edge detection in the image processing. C. Matching with database Fisher linear discriminant method is used match with the database if the features are matching then finally we can give the authenticated output. Fisher linear discriminant searches for projected vectors that best discriminate different classes in terms of maximizing the ratio of between-class to within class scatter. D. Locate the iris in color image In the color image process also segmentation will take place to locate the inner and outer boundary of iris using sobel filter to calculate the intensity value then using canny edge filter we should locate the inner boundary of iris in the color image and then huge variation in the intensity value from that we can locate the outer boundary of the iris from the eye color image, then we could convert the Cartesian form into polar form now the located iris from the eye color image is ready to extract the features[4][5].

The system is having basic structure is shown below the fig. 3 initially we should acquired the image then image is process in two various flow of recognition first flow is in gray scale image at the same time second flow is in color image then finally match the result of both the color image as well as gray scale how color image is accurate.

Fig. 2. Cartesian form into polar form

Fig. 3. With contact lens image

E. Feature Extraction for color image From the color image also we can extract the features like color, texture and shape for that we have used several algorithm to extract the features. For color image histogram technique is used to identify the value for the color image [6].For edge detection using canny edge detector and Euclidean Distance is used to identify the edge for the color image process Read in a color image and divide it into its three separate color channels. Run each color channel through the canny edge detector separately to find a resulting colored edge map. Combine the resulting edge maps from each of the three color channels into one complete edge. Edge detector should be able to detect edges in regions with high color variation but low intensity variation [8].The Euclidean Distance is a good operator for finding edges based on intensity. The distance between two pixels

F. Matching with database The feature is compare with the database to give the accurate result and then finally display the result whether they are authenticated or not for the matching algorithm Euclidean Distance is a good operator for finding edges based on intensity is used to compare the images. G. Comparison of gray as well as color images Comparing the number of users has been matched in gray scale as well as color image from that we can justify color image process is better than gray scale process. The basic procedure followed for contact lens problem can be overcome by using the above basic structure in fig. 4. In this module capturing of color image and locate the iris then identify the liveness of the iris image then feature extraction finally matching process will be processed. The basic with and without contact lens images are shown below the figure fig. 3 and fig. 5

Fig. 4. Contact Lens basic structure

H. Iris Edge Sharpness

Fig. 5. Without contact lens image

From the above two samples of with and without contact lens we can confine that there is a reflection of light from the contact lens approach is possible.

Fig. 7. Normal eye

Fig. 8. Affected eye

The image enhancement is having jobson Et Als Retinex [9]. These are the image enhancement technique are available. 3. IMPLEMENTATION

Fig. 6. Disease affected eye problem basic structure

The proposed system is having several advantages over the existing system. The system having totally 117 eye images of both the right as well as left eye image in color. The same color image is converted into gray scale and process the above algorithm for both gray scale image as well as color image the result of both the process are shown in below fig. 9(a). fig. 9(b).

The disease affected eyes usually several problems in eye to affected the like but iris wont affect frequently but when iritis and iridocyclitis these disease which affect the iris. The following figure shows the both normal eye and disease affected eye is shown in fig. 7 and fig. 8

Fig. 9(a). matching with database

Colorim e ag
120 100 80 60 40 20 0 Color image

Fig. 9(b). Not matching with database

The graph for both color image as well as gray scale image is plotted for authenticated and unauthenticated the graph shown below fig. 10 and fig. 11

Fig. 11. Color image graph 4. CONCLUSION

G S ray caleim e ag
100 90 80 70 60 50 40 30 20 10 0

The iris recognition is one of the best security system in present technology to give more strength to that technology using color image processing in iris recognition give very effective result when compare to the gray scale processing in iris recognition from the result itself it is justify. 5. FUTURE ENHANCEMENT

GrayScale im age

To make huge database and reduce the comparison time and to avoid contact lens problem as well as disease affected eye by using color image processing in iris recognition.
REFERENCES [1] [2] Iris Recognition System Using Fractal Dimensions of Haar Patterns Patnala S. R. Chandra Murty1, E. Sreenivasa Reddy2, and I. Ramesh Babu3 Research Scholar1, Professor2,3 Department of Computer Science and Engineering, Jawaharlal Technological University1,Vasireddy Venkatadri Institute of Technology2, Acharya Nagarjuna University3 Guntur, A.P., India. Iris detection using intensity and edge information Tsuyoshi Kawaguchi, Mohamed Rizon Department of Computer Science and Intelligent Systems, Oita University, 700 Dannoharu, Oita 8701192, Japan Received 3 May 2001; accepted 12 March 2002. Iris Recognition using Corner Detection P. Gupta1, H. Mehrotra1, A. Rattani1, A. Chatterjee1, A.K. Kaushik2 1 Indian Institute of Technology Kanpur, India


Fig. 10 Gray Scale image Graph



[6] [7] [8]


Department of Information Technology, New Delhi, India. Person identification technique using human iris recognition, Christel-loic Tisse, Lionel Martin, Lionel Torres, Michel Robert Advanced System Technology STMicroelectronics ZI Rousset 13106 Rousset Cedex, France. color image enhancement using single-scale retinex based on an improved image formation model doo hyun choi1, ick hoon jang2, mi hye kim3, and nam chul kim3.