The Fingerprint Recognizer Using Visual Cryptography

Published Date: 02 Nov 2017

Abstract: Fingerprint recognizer identifies the individual fingerprint from the fingerprint storage database. It will identify the fingerprint on the basis of count of Minutiae points found on the both matching fingerprint with the target fingerprint. Every human being has different structure of the fingerprints consists of the ridges and the valleys. Fingerprint recognizer takes the input as a fingerprint and applies the Visual Cryptography on it. After applying the Visual Cryptography, it will create the n number of transparencies of the fingerprint image. Here creating the two transparencies and storing them on two different servers. In the matching phase it will find Minutiae point locations on the fingerprint area, count the Minutiae points and store them on the features file. The features file will be referred during the target image will be match with the stored database. The system was developed using the without pixel expansion technique. This is the future work of this system.

Key terms

Fingerprint, Visual Cryptography, Minutiae Point, Transparency

Introduction

Visual Cryptography is the process in which Visual Cryptography algorithm will be applied on the input finger-print image. The algorithm creates the number of transparencies of the original fingerprint image. The given will be generate the two transparencies. Both transparencies will be stored on the two different image database servers. The system also provides the security to the image database servers. Existing systems based on the pixel expansion technique in which creates the two sub pixels from the original pixel. So the size of the original fingerprint image becomes the double in output image. Here without pixel expansion technique is used to identify the original fingerprint from the image database. In the without pixel expansion technique rotating the original fingerprint image by different angles such as 90,180 and 270 degree respectively without doing the pixel expansions. In encryption process original image pixels will be rotating the alternately using 4x4 block of pixel rotate by 90 degree, second by 180 degree and third by 270 degree again forth by 90 degree and so on. Then convert 2x2 blocks into 2x1 blocks. If more than three pixels are black then replace block with black pixel else replace with white pixel and name that image as "TEMPMS2".Then create random image "RS" by selecting the random pixels. Then created

image"MS2" by shifting the pixels of image"TEMPMS2"

using pixel values of"RS" In decryption process getting

the pixels of"MS2" and"RS" images from encryptions. Then shifting the pixels of "MS2" image using pixel values of "RS" image and named it as "TEMPORIGINAL". From the"TEMPORIGINAL" image convert block of 2x1 pixels into 2x2 pixels. If 2 pixels of block are black then replace it with black pixel block else replace it with white pixel and

named it as"ORIGINAL" image. Then done the reverse process of encryption rotate the alternate blocks of" ORIGINAL" image by different angles to get the decrypted image.

Related Work

Visual Cryptography id the technique in which the original fingerprint image was divides in to the number of transparencies. Each transparency will be stored on the separate image database server. Here creating the two transparencies of the original fingerprint image. Naor and Shamir[2] have already developed the algorithms on the Visual cryptography. They invented the Visual Cryptography in 1994. To secure the Arun Ross and Asem Othman[1] also works on the Visual Cryptography. They secure the biometric database using the pixel expansion technique. Here using the without pixel expansion technique to identify and protect the fingerprint images stored on the image database using the Visual Cryptography.

In without pixel expansion technique rotating the original fingerprint image by different angles, such as 90 degree, 180 degree, 270 degree and so on. Ching-lin Wang, Ching-Te Wang and Meng-Lin Chiang [15] present the Image Multiple Secret Sharing Schemes Without pixel expansion in that they take the multiple source images and create the two share images. In this system taking the single source image and rotate it by different angles to create the two transparency images.

To protect the biometric database using visual cryptography technique with pixel expansion makes the size of the original finger-print image double presented by the Arun Ross and Asem Othman [1].To assure that ,the new technique of without pixel expansion makes the size of the template image 50 percent less than the original fingerprint image. In this project implemented the single secret sharing scheme without pixel expansion. Using the with-out pixel expansion technique Original fingerprint image are rotate in different angles to create the encrypted image using the 4X4 block of the fingerprint image.

Drawbacks of existing systems:

Due to the pixel expansion method size of original image becomes double.

Due to the pixel expansion, difficult to find out exact Minutiae points.

Fingerprint matching process becomes the difficult.

M’= set of Minutiae points of the Target fingerprint image

Therefore, M’={M1’,M2’,M3’,…..Mn’}

M’ ≡ M, both sets are equivalents then Target fingerprint image will be match with the Original fingerprint image.

Data Flow Architecture

The data flow diagram depicts the information flow and transformations that are applied on the data as data moves from input to output. Figure 1 shows the working flow of the system using the tree diagram as shown in the figure 1.

Programmer’s design

Fingerprint recognizer works on the basis of without pixel expansion method. It takes the fingerprint image as in-put. After getting the input it will apply the Visual Cryptography to create the two transparencies from the original fingerprint image.

Mathematical Model

Let, I= is a original input fingerprint image......Equation (1)

We are dividing original fingerprint image into two transparencies and store on the two different servers.

Let,

S be set of

transparencies from server1 and

S’ be the set of transparencies from the server2

Set

Elements

Semantics

I

I1,I2,...In

Set of fingerprint images

S

S1,S2,...Sn

Set of transparencies on server1

S0

S1,S2,...Sn

Set of transparencies on server2

M

M1,M2,...Mn

Set of Minutiae points

M0

M1,M2,...Mn

Set of Minutiae points of target

image

Table 1: Mathematical set For Fingerprint Recognizer

S= S1, S2, S3...Sn......Equation (2)

S’ = S1, S2, S3 ...Sn......Equation (3) Figure 1: Data flow Architecture

From Equation (1), (2) and (3)

S ε I and S’ε I

where S and S’ both transparencies are

belongs to the Original fingerprint image I

S Є I and S’ Є I………both transparencies are belongs to the Original fingerprint image I

So that, I= S U S’

Let, M be set of Minutiae points of the Original fingerprint image I stored on the database

Figure 2: Level1 DFD

Turing Machine

Figure3 shows the system state diagram. In that sys-tem will take the fingerprint image as input and apply the visual cryptography to create the two transparencies. After creating the transparencies, store on the two different databases. At the time of Minutiae point finding both transparencies are combined together.

Figure 3: System state diagram

Results

Figure3 and figure4 shows the outcome transparencies of the original fingerprint image after applying the Visual Cryptography techniques. Transparencies are created using the Visual Cryptography in pixels from the original fingerprint image are chosen randomly using the random generator.Fig.5 Shows the Minutiae points in those green dots represent the Minutiae points. Minutiae points can be find using the minutiae extraction algorithm which finds the locations of the Minutiae points. Matching can be done on the basis of number of Minutiae points. If the required Minutiae count will be match with the stored fingerprint image Minutiae then it will display the match found result else match not found. Figure3 is

Figure 4: Transparency1

the first output transparency image after applying the Visual

Cryptography.

Figure 5: Transparency2

Figure 6: Minutiae points

Conclusion

The objective behind designing this system is to design the fingerprint recognizer using the without pixel expansion technique which is better one than the existing pixel expansion method. So, finally implemented the fingerprint recognizer using the without pixel expansion method.