2011 Fourth International Conference on Intelligent Computation Technology and Automation

The design and implementation of ID Authentication System Based on Fingerprint Identification

Zhang Jinhai, Liu Xinjian, Chen Bo
Marine college of Shandong jiaotong university, Weihai , Shandong, 264200,China zhangjinhai76@163.com

AbstractAutomatic identification technology becomes a urgent need of production and life, authentication technology gained worldwide attention because of its high reliability, fingerprint identification technology which applied to social security system can accurately determine protects a person's identity and prevent the phenomenon of the pension of falsely claim that solves this one long-term puzzling problem. Based on the preeminent fingerprint identification algorithm is improved, and combining with actual demand the improved algorithm, Using the fingerprint identification and IC card combination way realization of distributed fingerprint authentication system collection. Through the practical efficiency test analysis proves the whole system is feasible, and the running effect is good by practical application.
Keywords- fingerprint identification; Component technology; IC card applications

planning, and finally puts forward some fingerprint identification authentication center of the whole system scheme. II.
THE PRINCIPLE OF FINGERPRINT IDENTIFICATION

I. INTRODUCTION Since the middle of 1980s, with annuities socialization extends the all-round development (national enterprise basic annuities socialization extends toll 33.746 million people, already executes socialization management services have 18.302 million), pension falsely claim problem has become more obvious. Preliminary estimates, the total number of falsely claim issue of the total number of 3-5 per 1,000, some regions and even up to 5%, causing the loss of national total annual pension up to a billion Yuan. For block pension loss of "black hole", all social security organization adopted various measures, such as: the specified movement photograph, with monthly newspapers, magazines, photo, local police station to issue survival certificate, etc. These methods in the actual operation process cannot repeat use, big workload, personnel opinion bigger also. Anyhow, the traditional ways of falsely claim prevention existing many disadvantages. Due to human body features with no duplicated characteristics, people put eyes turned to biometrics. Computer fingerprint identification automatic identification technology is the earliest and the most practical and the mature technology among the identification technology. through the detailed analysis for automatic fingerprint identification system needs, and according to the system's demand for fingerprint identification system for the overall analysis and design, first fingerprint collection and processing scheme design and implementation, and then the fingerprint information processing to carry on the detailed
978-0-7695-4353-6/11 $26.00 2011 IEEE DOI 10.1109/ICICTA.2011.585 1234 1217

Automatic fingerprint Identification system work mode can be divided into two types: Verification mode and Identification mode. Verification mode also called one-toone ratio (l: 1), principle for Matching: fingerprint preregister to sample database and set a logo yards, when matching, inputting fingerprint and input identification code, system identification code from database according to extracting fingerprint sample and own fingerprint compares. Identification mode which also called a pair of dobby (l: N Matching), is to treat entry fingerprints and sample database of all fingerprints were compared, until find one match fingerprints or search the entire sample database are no corresponding fingerprints after the conclusion. Automatic fingerprint identification system mainly involves image collection, pretreatment, feature extraction, and feature matching and so on several parts. The system consists of image input, pretreatment, feature extraction, classification and matching parts, among these five parts pretreatment part can divide again for image segmentation, image enhancement, binary conversion and refining etc. III.
THE FINGERPRINT IDENTIFICATION ALGORITHMS.

Fingerprint identification include the following steps: fingerprint image pretreatment, fingerprint characteristic extraction, fingerprint characteristic comparison, a typical fingerprint identification algorithm process as shown in figure 1 shows:

Figure 1 a typical fingerprint identification algorithm process

The algorithm for the fingerprint of image processing uses the many kinds of optimization algorithm, including the frequency domain and time domain two ways of combining the filtering processing technology which be able to effectively filter fingerprint noise. At the same time its

unique fingerprint quality (such as dry humidity) evaluation factors, can effective adjustment the quality of pixels of normal direction. For dryness and broken mark fingerprints, by analyzing the fingerprint direction can effectively compensate amount. Total processing steps can reach 17 steps above. The image processing of simplified as figure 2 shows:

IMPLEMENTATION OF FINGERPRINT IDENTIFICATION SYSTEM

The fingerprint identification system is designed in strict accordance with the labor bureau work requirements and safety guidelines, generality, portability, more information and higher safety requirements of the system are full consideration. The fingerprint collection and fingerprint verification are the most main activity. A. implementation of fingerprint acquisition Maintaining the Integrity of the Specifications Fingerprint acquisition subsystem in the program uses PowerBuilder compile which makes the fingerprint instrument operation and fingerprint identification algorithm encapsulation into DLL and proceeds necessary function declaration. Fingerprint characteristic information table in table 1.
Table 1 fingerprint characteristic information table

Original figure after image enhancement after smooth processing after binaryzation after thinning Figure 2 simplified diagram of image processing process

IV.

ANALYSIS AND DESIGN OF THE FINGERPRINT IDENTIFICATION SYSTEM

Item
EMPNO Fingerprint NO Feature point Feature Information Fingerprint Image

Field Name
UserId FingerId MunutiaeNu m MunutiaeInfo FingerImg

datat ype
VARC HAR2 Int Int Long Binary Long Binary

Data Length
10

Nullable
NOTNULL NOTNULL NULL NULL NULL

Fingerprint identification authentication center system composes of the database part, system management section, application server part, user management workstations and business management workstation five major components. Attend insurance personnel, system administrators, business personnel, information of attend insurance personnel are the most basic object class, the main operations are attend insurance personnel information registration, fingerprint acquisition, fingerprint verification, validation information query. The most main activity is the fingerprint collection and fingerprint verification. Social security card information collection service system strives to construct an electronic and effective, synthetic multi-function graphic information management and survival verification system, annuities receivers survive verification validation are judged through fingerprint identification technology as the main technical means, supplemented by other graphic information (signature, photo, ID, bank account passbook, etc.) verification methods, which makes the social security organization endowment insurance business system more perfect. Social security card information collection service system includes the following function module: attend insurance personnel file, the social security core platform. The core platform is the core part of the whole social insurance system, which including endowment annuity, unemployment insurance, medical insurance, insurance against injury at work and birth insurance by business divisions and each subsystem are both run separately or combined together operation arbitrarily. Data sources of social insurance management information system include attend insurance unit, bank, taxation, designated medical establishment retail medicamentarius information system, the core platform has also data interface with these external system, so that handing data exchange.

B. implementation of fingerprint authentication Fingerprint identification authentication subsystem makes fingerprint identification algorithm encapsulation into DLL using PowerBuilder compile and in PowerBuilder in necessary function declarations, then according to the following processes realize fingerprint identity authentication, fig. 3 shows.

Figure 3 fingerprint authentication process

1218 1235

V.

IMPLEMENTATION OF IC CARD DATA ACCESS

The IC card issuing subsystem, written by the insured person after the basic information and fingerprint information, IC cards have been preserved in the hands of the insured, there must be some measures to ensure the IC card can not be illegal to modify the situation create counterfeit cards. The IC card chip itself has a write protection feature, you must first verify the information before writing IC card password (of course change this password must also verify the old password), which can guarantee a certain extent, not be illegal to write data. But still can not completely avoid this from happening (such as write card programs are illegally obtained), it is necessary to verify the validity of the information read. At this stage, there are two common encryption method, one-way encryption and two-way encryption. Two-way encryption is the most commonly used encryption algorithm, it's clear that people can understand the data encryption is not directly understand the ciphertext data, and then, when needed, you can use some of these encryption algorithm to decrypt the ciphertext after can understand the original plaintext. Two-way encryption for secret communications, for example, online shopping, credit card need to submit your site, people do not want to send clear text data directly online, as this is likely to be another user "eavesdropping" People want credit card After through encryption and then transmitted in the network, so that after the site receives data through the decryption algorithm can be accurate to the credit card account. The system used MD5 (Message-Digest algorithm 5) one-way encryption algorithm, which consisted of MD2, MD3, MD4 developed from a one-way function algorithm (that was, HASH algorithm), which was an internationally renowned public-key encryption algorithm standard RSA, which developed by the first designer R.Rivest in the early 90s of last century. The most important function of MD5, the large files of different formatted of information in the software used to sign digital signature private key before the "compressed" into a confidential format, the key is that this "compression" is not reversible. For MD5, there are two characteristics are very important, first, any two plaintext data, cipher text after encryption can not be the same, the second is a clear text any data after the encrypted, the result must always be unchanged. MD5 has a 128-bit integer variable called the link parameters, as long as they have some modifications to the equivalent of a different MD5 algorithm. In this system, IC card information by checking MD5 checksum to determine the legality of the information. Write the first stage of the IC card and fingerprint information according to characteristics of insured value of the information string, resulting in MD5 code, and write to the IC card. MD5 interest code, and write to the IC card. In the IC card reader, you need to set, read the information section of the IC card to generate MD5 code and IC card and compare the MD5 code to determine the legality of the information. MD5 code as each generation is changing the initial value (by social security center managers control),

so even if the IC card for the illegal modification of information written procedures, by making it impossible to verify the information when the legitimacy of validation. IC card reader on the general process has been improved to achieve the same procedure used in the preparation of PowerBuilder, Dynamic Link Library by calling the IC card reading and writing. A. Fingerprint Authentication Platform The system developed the fingerprint authentication platform, based on information security, biometric fingerprint all fully focused on the platform database control, which must all access through the platform, not stored in the user side computer. Figure 4 shows.

Figure 4 the system platform

The system developed by the fingerprint authentication platform used a distributed architecture, can be divided into two major areas: server and client, server, save the main provider of fingerprint information, fingerprint authentication and enterprise applications and other functions, and the client provides fingerprint capture and enterprise applications users use interface, which detailed and packaged on the Following block. Fingerprint recognized system architecture platform shown in Figure 5.
Server XServer

XAgent Client XEnroll XClient XVerify Fingerprint Driver

Figure 5 the system architecture of platform

B. System Performance Tested In order to test the reliability and practicality of the system conducted a number of small-scale tested of

1219 1236

fingerprint recognition applications. The test results which can be seen from Table 2
Table 2 test results

Test time 2008.4.3 2008.4.5 2008.4.8 2008.4.9 2008.4.11 2008.4.12

Person participated 600 600 600 600 600 600

The rated of average passed 98.3% 98.6% 99.3% 99.7% 100% 100%

[9] A. Bishmu, S. Das, and S. Nandy. An improved Algorithm for Point Set Pattern Matehing under Rigid Motion.CIAC,LNCS 2653. 2003:36~45 [10] N. Alinik, R. Atha, and K. Allek, ect. A real-time matching system for large fingerprint database[J]. IEEE Trans on Patern Analysis and Machine intelligence, 1996,18(8):799~812

Table statistics showed that application of fingerprint identification test pass rates over time showed a significant increasing trend, mainly due to: first use of fingerprint authentication, staff and the insured are relatively tight, the operation was not standardized, unskilled. With the increased use of proficiency and the gradually stable moods of insurance person, the second validation pass rate rising rapidly. After the first two machine running, the third time after more than 99% pass rate, fundamental to the expected normal values. VI. CONCLUSION Fingerprint authentication system applied to the social endowment insurance system, has basically met the needs of the first set, can be achieved, or for retirees who retired fingerprint registration formalities, and then retired regular or irregular acquisition of the original fingerprint to compare the fingerprint to verify the retirees to the survival of the state's purposes. Fingerprint Identification System fingerprint authentication state and status of real-time feedback to the business system. The pensions system in the real alternative to the traditional manual identification of the identity of the insured's work, improve efficiency, effectively prevent the occurrence impersonator pensions, more easily verify the conditions of the late management. REFERENCES
[1] Tianjie,Yang xin. Theory and application of Biometrics identification technology. Publishing House of Electronics Industry..2006:56-77 [2] Wang shengguo. Fingerprint recognition algorithm research and realization implement. China excellent master degree theses full-text database. Hainan normal university. 2007:38~41 [3] Ren chunxiao. Fingerprint algorithm study platform design development and application. Shandong university. 2007:66~79 [4] Liang chuan. The Scenario and Application of Fingerprint Authentication with Electronic Request-Authorization System in Traffic Business. Sichuan university. 2007,(4) [5]A. Cavoukian, J. Stoianov. A Positive-Sum Technology that Achieves Strong. Authentication, Security and Privac. 2007,(3) [6]D. Simon, Z. Jortega, S. Llanas and J. Gonzalez. Minutiae Extractios chemefor fingerprint recognition systems. International Conferenceon Image Processing(ICIP),Greece, 2001,10:7~10 [7]W. Green, D. John. Internet and Distributed Application Development Beijing mechanical industry press.. 2004(7) [8] K. J. Anil, P. Bolle. An Identity-Authentication System Using Fingerprints Proceedings of The IEEE. 1997,9(9):85

1220 1237