A Seminar Report On FINGERPRINT VERIFICATION SYSTEM

In partial fulfilment of requirements for the

Bachelors Degree in Computer Engineering

SUBMITTED BY:

PRAJAKTA MORE

Under the Guidance of

Prof. MATANG J. SURTI

DEPARTMENT OF COMPUTER ENGINEERING

Sandip Institute of Engineering & Management, Nasik

2012-13

ROLL NO: 37/T.E.

CERTIFICATE
This is to certify that the Seminar titled

FINGERPRINT VERIFICATION SYSTEM

was presented by

PRAJAKTA MORE
(Roll No.:37)

of the Sixth Semester Computer Engineering in partial fulfilment of requirement of Degree of Bachelor in Computer Engineering at Sandip Institute of Engineering & technology University of Pune during the year 2012-13

SEMINAR GUIDE

HEAD OF THE DEPARTMENT

Examiner

ACKNOWLEDGEMENT

I would like to articulate my deep gratitude to my seminar guide Prof.M.J.Surti for his advice, assistance and constant support in the seminar work. I would like to thank The Head of Department for his throughout guidance and would like to thank all faculty members and staff of the Department of Computer Engineering, S.I.E.M, Nasik for their generous help in various ways for this seminar. Last but not the least, my sincere thanks to all of my friends who have patiently extended all sorts of help in this seminar.

ABSTRACT

Fingerprint recognition is one of the most popular and accuracy Biometric technologies. Nowadays, it is used in many real Applications. In recent years, many algorithms, models are given to improve the accuracy of Recognition system. In this model, after pre-processing step, we find the transformation between templates, adjust parameters, synthesize fingerprint and reduce noises. Then, we use the final fingerprint to match with others in FVC2004 fingerprint database (DB4) to show the capability of the model. In fingerprint recognition process, the important step which affects on system accuracy is matching between template and query fingerprint. Fingerprint recognition is done using minutiae extraction in time domain method and using Discrete Cosine Transform (DCT) in frequency domain method. Human fingerprints are rich in details which is known as minutiae, which can be used as identification marks for fingerprint verification. The aim of this report is to study on fingerprint recognition system based on minutiae based matching which is quiet frequently used in various fingerprint algorithms and techniques. The approach of this report involves how the minutiae points are extracted from the fingerprint images and after that between two fingerprints perform the fingerprint matching. Image enhancement, image segmentation, minutiae extraction and minutiae matching these stages are the main themes of the seminar.

CONTENTS
Page No Certificate.2 Acknowledgement3 Abstract4 List if tables..7 List of figures...7 CHAPTER 1 INTRODUCTION 1 Introduction..8 1.1Fingerprint..9 1.2Applications9 1.3Advantages...10 1.4Limitations10 CHAPTER 2 LITERATURE SURVEY 2.1Techniques for Fingerprint Matching...............................................................12 2.2Fingerprint Identification.................................................................................13 2.2.1Minutiae Extraction..................................................................................15 2.2.2 Minutiae Matching...................................................................................19 2.2.3Fingerprint sensors....................................................................................22

2.1.3 CHAPTER 3 METHODOLOGY CHAPTER 4 ALGORITHM CHAPTER 5 RESULT ANALYSIS CHAPTER 6 FUTURE SCOPE

CHAPTER 7 CONCLUSION

REFERENCES

LIST OF TABLES
Page No Table 1.3.1 Typical applications of biometrics for positive person identification...........................................................................................12

LIST OF FIGURES
Page No Fig 1.1 Fingerprint, Hand and Iris system Input.8 Fig 2.1.1 Fingerprint Identification....................................................................13 Fig 2.1.2 Block Diagram of Fingerprint Scanner...14 Fig 2.1.3 Classification of fingerprints on the basis of geometric structure...15 Fig 2.1.3 Fingerprint Ridge Thinning.................................................................16 Fig 2.1.4 Fingerprint minutiae marking..16 Fig 2.1.5 Three step approach of minutiae extraction..17 Fig 2.1.6 Images correctly detected as a match by the matching algorithm...19 Fig 2.2.1 Minutiae Matcher..................................................................................19 Fig.2.2.6 One by one steps involved in fingerprint recognition algorithm.........21 Fig 2.3.1 Finger print image acquired by a Sensor.............................................22 Fig 5 Fig 6 Fig 7 Fig 8 Fig 9 Fig 10

Chapter 1 Introduction
1. Introduction More than a century has passed since Alphonse Bertillon rst conceived and then industriously practiced the idea of using body measurements for solving crimes. The scientific literature on quantitative measurement of humans for the purpose of identification dates back to the 1870s. Bertillons system of body measurements, including such measures as skull diameter and arm and foot length, was used in the USA to identify prisoners until the 1920s.Soon after this discovery, many major law enforcement departments embraced the idea of rst booking the ngerprints of criminals, so that their records are readily available and later using leftover ngerprint smudges (latents), the identity of criminals can be determined. Basically skin of human fingerprints consist of ridges and valleys and they mixing together form the distinctive patterns. At the time of pregnancy these distinctive patterns are fully developed and are permanent throughout the whole lifespan. Those patterns are called Fingerprints. From observations we conclude that fingerprints offer more secure and reliable personal identification than passwords or id-cards. Immigration cards holding both passport number and measures of the users hand; fingerprints taken as a legal requirement for a driver license, but not stored anywhere on the license; season tickets to an amusement park linked to the shape of the purchasers fingers; and confidential delivery of health care through iris recognition; these systems seem completely different in terms of purpose, procedures, and technologies to identify a human uniquely.

Fig.1 Fingerprint, hand and iris system input images

1.1 Fingerprint Basically Skin of human fingertips consists of ridges and valleys and they mixing together form the distinctive patterns. Those patterns are called fingerprints. From different researches it has been observed that no two persons have the same fingerprints. A ridge is defined as a single curved segment, and a valley is the region between two adjacent ridges. Minutiae points are the local ridge discontinuities, which are of two types: ridge endings and bifurcations. A good quality image has around 40 to 100 minutiae. It is these minutiae points which are used for determining uniqueness of a fingerprint. Fingerprints are imprints formed by friction ridges of the skin and thumbs. They have long been used for identification because of their immutability and individuality. Immutability refers to the permanent and unchanging character of the pattern on each finger. Individuality refers to the uniqueness of ridge details across individuals; the probability that two fingerprints are alike is about 1 in 15 1.9x10 . Each individual has its own fingerprint with permanent uniqueness. Thats why fingerprints have been used for identification and forensic investigation for a long time. The main parameters characterizing a fingerprint image are Resolution Area Number of pixels Dynamic Range Geometric Accuracy Image Quality

1.2 Applications As mentioned earlier, law enforcement agencies were the earliest adopters of the fingerprint identification technology. More recently, increasing identity fraud has created a growing need for biometric technology for positive person identification in a number of non-forensic applications. Is this person authorized to enter this facility? Is this individual entitled to access the privileged information? Is the given service being administered exclusively to the enrolled users? Answers to questions such as these are valuable to business and government organizations. Since biometric identifiers cannot be easily misplaced, forged, or shared, they are considered more reliable for personal identification than traditional token or knowledge based methods.

Table 1.3.1 Typical applications of biometrics for positive person identification.

Forensic

Civilian

Commercial ATM Access control Cellular phone Credit card

Corpse identication National ID Criminal Investigation Drivers license Parenthood determination Welfare disbursement Border crossing

1.3 Advantages Fingerprints do not change over time. Fingerprints stop unauthorized access. All fingers are unique, which allows each person to have ten easy uses of identifiers. Base of all world-wide identification. Fast and easy to use. Users respect them, fraudsters are afraid of them because of its feasibility, distinctiveness, permanence, accuracy, reliability, and acceptability. Its maturity, providing a high level of recognition accuracy. The growing market of low-cost small-size acquisition devices, allowing its use in a broad range of applications, e.g., electronic commerce, physical access, PC logon, etc. The use of easy-to-use, ergonomic devices, not requiring complex user-system interaction.

1.4 Limitations While significant progress has been made in automatic fingerprint identification, there are still a number of research issues that need to be addressed to improve system accuracy. Most of the shortcomings in the accuracy of an automatic fingerprint identification system can be attributed to the acquisition process.

10

(i) Inconsistent contact: The act of sensing distorts the fingerprint. Determined by the pressure and contact of the finger on the glass platen, the three-dimensional shape of the finger gets mapped onto the two-dimensional surface of the glass platen. As the finger is not a rigid object and since the process of projecting the finger surface onto the image acquisition surface is not precisely controlled, different impressions of a finger are related to each other by various transformations. The most problematic of these projections appears to be elastic distortions of the friction skin of finger which displaces different portions of the finger (ever so slightly) by different magnitudes and in different directions. (iii) Non-uniform contact: The ridge structure of a finger would be completely captured if ridges belonging to the part of the finger being imaged are in complete physical/optical contact with the image acquisition surface and the valleys do not make any contact with the image acquisition surface. However, the dryness of the skin, shallow/worn-out ridges (due to aging/genetics), skin disease, sweat, dirt, humidity in the air all confound the situation resulting in a non-ideal contact situation. In the case of inked fingerprints, an additional factor may include inappropriate inking of the finger: This results in noisy low contrast images, which leads to either spurious or missing minutiae. (iv) Irreproducible contact: Manual work, accidents, etc. inflict injuries to the finger, thereby, changing the ridge structure of the finger either permanently or semi-permanently. Further, each impression of a finger may possibly depict a different portion of its surface. This may introduce additional spurious fingerprint features. (v)Feature extraction artifacts: Various image processing operations might introduce inconsistent biases to perturb the location and orientation estimates of the reported fingerprint structures from their grayscale counterparts. (vi) The act of sensing itself adds noise to the image. For example, in the livescan fingerprint acquisition method, residues from the previous fingerprint capture may be left behind. A typical imaging system geometrically distorts the image of the object being sensed due to imperfect imaging conditions. (vii) Factors such as finger injuries or manual working, can result in certain users being unable to use a fingerprint-based recognition system, either temporarily or permanently. (viii)Small-area sensors embedded in portable devices may result in less information available from a fingerprint and/or little overlap between different acquisitions.

11

CHAPTER 2 LITERATURE SURVEY

2.1 Techniques for Fingerprint Matching (1)Minutiae-based technique: In this, the terminations and bifurcations of the ridge lines attempt to align two sets of minutiae points and determine the total number of matched minutiae. Minutiae based techniques represent the fingerprint by its local features, like terminations and bifurcations. Two fingerprints match if their minutiae points match. This approach has been intensively studied, also is the backbone of the current available fingerprint recognition products. (2)Pattern Matching or Ridge Feature Based Technique: Feature extraction and template generation are based on series of ridges as opposed to discrete points which forms the basis of Pattern Matching Techniques. This includes context aware similarity search techniques applicable to all types of content based image retrieval (CBIR). The advantage of Pattern Matching techniques over Minutiae based techniques is that minutiae points may be affected by wear and tear and the disadvantages are that these are sensitive to proper placement of finger and need large storage for templates. (3)Correlation Based Technique: Let I(x, y, ) represent a rotation of the input image I by an angle around the origin (usually the image center) and shifted by x and y pixels in directions x and y, respectively. Then the similarity between the two fingerprint images T and I can be measured as :

S(T,I)=
T

where CC(T, I) = T I is the cross-correlation between T and I. The crosscorrelation is a well known measure of image similarity. It allows us to find the optimal registration. The direct application rarely leads to acceptable results, mainly due to the following problems: a) Non-linear distortion makes impressions of the same finger significantly different in terms of global structure; the use of local or block-wise correlation techniques can help to deal with this problem.

12

b) Skin condition and finger pressure cause image brightness, contrast, and ridge thickness to vary significantly across different impressions. The use of more sophisticated correlation measures may compensate for these problems. c) The technique is computationally very expensive. Local correlation and correlation in the Fourier domain can improve efficiency.

(4)Image Based Techniques: Image based techniques try to do matching based on the global features of a whole fingerprint image. It is an advanced and newly emerging method for fingerprint recognition. It is useful to solve some intractable problems of the first approach.

2.2 Fingerprint Identification Automated fingerprint recognition and self authentication system can be categorized as verification or identification systems. Fingerprint verification is the method where we compare a claimant fingerprint with an enrolee fingerprint, where our aim is to match both the fingerprints. This method is mainly used to verify a persons authenticity. For verification a person needs to his or her fingerprint in to the fingerprint verification system. Then the representation is saved in some compress format with the persons identity and his or her name. Then it is applied to the fingerprint verification system so that the persons identity can be easily verified. Fingerprint verification is also called, one-to-one matching.

Fig 2.1.1 Fingerprint Identification

13

A fingerprint identification system is an automatic pattern recognition system which consists of the three fundamental stages: (i) (ii) (iii) data acquisition : the fingerprint to be recognized is sensed; feature extraction : a machine representation (pattern) is extracted from the sensed image; and decision-making : the representations derived from the sensed image are compared with a representation stored in the system.

The fingerprint recognition problem can be grouped into three sub-domains: fingerprint enrollment, verification and fingerprint identification. In addition, as different from the manual approach for fingerprint recognition by experts, the fingerprint recognition here is referred as AFRS (Automatic Fingerprint Recognition System), which is program-based. Verification is typically used for positive recognition, where the aim is to prevent multiple people from using the same identity. Fingerprint verification is to verify the authenticity of one person by his fingerprint. There is one-to-one comparison in this case. In the identification mode, the system recognizes an individual by searching the templates of all the users in the database for a match. Therefore, the system conducts a one to many comparisons to establish an individuals identity. Both verification and identification use certain techniques for fingerprint matching as indicated in the following subsection. The verification process either accepts or rejects the users identity by matching against an existing fingerprint database. In identification, the identity of the user is established using fingerprints. Since accurate matching of fingerprints depends largely on ridge structures, the quality of the fingerprint image is of critical importance. However, in practice, a fingerprint image may not always be well defined due to elements of noise that corrupt the clarity of the ridge structures. This corruption may occur due to variations in skin and impression conditions such as scars, humidity, dirt, and non-uniform contact with the fingerprint capture device.

14

Fig 2.1.2 Block Diagram of Fingerprint Scanner

Some important terms related to fingerprint identification systems are explained below : Fingerprint Acquisition: How to acquire fingerprint images and how to represent them in a proper machine-readable format. Depending on whether the acquisition process is offline or online, a fingerprint may be either (i) an inked fingerprint, (ii) a latent fingerprint or (iii) a live-scan fingerprint. Fingerprint Verification: To determine whether two fingerprints are from the same finger. Fingerprint Identification: To search for a query fingerprint in a database. Fingerprint Classification: To assign a given fingerprint to one of the prespecified categories according to its geometric characteristics.

In case of both fingerprint identification and fingerprint verification systems, the tasks will be broken up into 2 stages: a. Off-line phase: Several fingerprint images of the fingerprint of a person to be verified are first captured and processed by a feature extraction module; the extracted features are stored as templates in a database for later use.

15

b. On-line phase: The individual to be verified gives his/her identity (in case of a verification system) and places his/her finger on the inkless fingerprint scanner, minutiae points are extracted from the captured fingerprint image. These minutiae are then fed to a matching module, which matches them against his/her own templates in the database (in case of a verification system) or against all the users in the database (in case of an identification system). In case of a fingerprint identification system, we would need to match the incoming fingerprint with stored templates of every other user in the database. In order to reduce this computation and search overhead, it is essential to have some kind of fingerprint classification system which will help us to severely restrict the size of the database for which we need to extract minutiae features and match against the incoming fingerprint sample. Hence, fingerprints can be classified into one of six categories based on their geometric structure at a macro level.

(a) arch

(b) tented arch

(c) right loop

(d) left loop

(e) whorl

(f) twin loop

Fig 2.1.3 Classification of fingerprints on the basis of geometric structure

Hence, by classifying fingerprints on the above basis we can make the database to be searched much smaller. Also, if images from multiple fingers are used, we will have many more categories and the size of the database to be searched will decrease further.

16

Fig 2.1.4 Block Diagram to test the fingerprint match

The fingerprint verification problem can be divided into following main tasks as shown in the above block diagram : 2.2.1 Image Aquisition

Fingerprint Image: The input fingerprint image is the gray scale image of a person, which has intensity values ranging from 0 to 255. In a fingerprint image, the ridges appear as dark lines while the valleys are the light areas between the ridges. Minutiae points are the locations where a ridge becomes discontinuous. A ridge can either come to an end, which is called as termination or it can split into

17

two ridges, which is called as bifurcation. The two minutiae types of terminations and bifurcations are of more interest for further processes compared to other features of a fingerprint image.

Fig 2.1.5 A fingerprint image with the core and four minutiae points labeled

2.2.2

Normalization

For normalization we have done simple histogram equalization, which enhances the contrast of images by transforming the values in the fingerprint image.

Fig 2.2.1 Normalisation

2.2.3 Ridge Orientation Implications of poor quality image: A significant number of spurious minutiae may be created, A large percent of genuine minutiae may be ignored, and

18

 Large errors in their localization (position and orientation) may be introduced. Thus, from the normalized fingerprint image orientation of ridges are calculated in each block of desired size. Then, using the normalized image and the Orientation image the frequency in each of the blocks of the image is calculated. 2.2.4 Filtering

A bank of Gabor filter tuned to local ridge orientation and frequency is used to filter the image to clearly separate out ridges and valleys and thus reduce the probability of spurious minutiae. The sinusoidal-shaped waves of ridges and valleys vary slowly in a local constant orientation. Therefore, a band-pass filter that is tuned to the corresponding frequency and orientation can efficiently remove the undesired noise and preserve the true ridge and valley structures. Gabor filters have both frequency-selective and orientation selective properties and have optimal joint resolution in both spatial and frequency domains. Therefore, it is appropriate to use Gabor filters as band-pass filters to remove the noise and preserve true ridge/valley structures.

Fig 2.3.1 Filtering

2.2.5

Thinning

Due to noise limitation on image acquisition, skin deformations etc the fingerprint image obtained is never ideal. As a result there are a lot of spurious minutiae that crop up if we simply follow the above approach to minutiae detection. To solve the problem, various heuristics have been proposed and we have implemented the following rules to remove most of the spurious minutiae, resulting from noise in the thinned image: If several minutiae form a cluster in a small region, then remove all of them except for the one nearest to the cluster center If two minutiae are located close enough, facing each other, but no ridges lie between them, then remove both of them In addition to the noise in the fingerprint image, the thinned image may not be ideal. If such is the case, minutiae extraction may not yield the correct results.

19

(a)Original image

(b) image after thinning

Fig 2.1.3 Fingerprint Ridge Thinning

2.2.6

Binarization

(a) Original Fingerprint Fig 2.6.1 Binarization

(b) Binarized image.

2.2.1 Minutiae Extraction Among all the fingerprint features, minutiae point features with corresponding orientation maps are unique enough to discriminate amongst fingerprints robustly; the minutiae feature representation reduces the complex fingerprint recognition problem to a point pattern matching problem. In order to achieve high-accuracy minutiae with varied quality fingerprint images, segmentation algorithm needs to separate foreground from noisy background which includes all ridge-valley regions and not the background. Image

20

enhancement algorithm needs to keep the original ridge flow pattern without altering the singularity, join broken ridges, clean artifacts between pseudo-parallel ridges, and not introduce false information. Finally minutiae detection algorithm needs to locate efficiently and accurately the minutiae points.
We used iterative parallel thinning algorithm for minutiae extraction stage. Ridge thinning is used to used to eliminate the redundant pixels of the ridges till the ridges are of one pixel wide. The minutiae marking is quite simple task. Here crossing number (CN) concept is used.

(a)Extracted minutiae

(b) Zoomed section of the image

Fig 2.1.4 Fingerprint minutiae marking

Minutiae extraction from a perfectly thinned ridge-map of a fingerprint image is a trivial task. All we need to do is to count the number of ridge pixels, every ridge pixel on the thinned image is surrounded by and depending on the following rule, and we can assign the minutiae points to those pixels: To extract a three minutiae a three step approach is used such as : i)pre processing stage ii)minutiae extraction stage iii)post processing stage

IMAGE ENHANCEMENT

PREPROCESSING
IMAGE BINARIZATION IMAGE SEGMENTATION

MINUTIAE EXTRACTION

segmentation THINNING

MINUTIAE MARKING

POSTPROCESSING 21

REMOVE FALSE MINUTIAE

REMOVE FALSE MINUTIAE

Fig 2.1.5 Three step approach of minutiae extraction

Pre processing stage:Again pre processing stage is divided in to three sub stages such as:- i) image enhancement ii) image binarization iii) image segmentation. For image enhancement we used two methods such as:- histogram equalization and Fourier transform. After enhancing the image we need to binaries the image for that we used the locally adaptive threshold method. For image segmentation we preferred a three-step approach such as :- i) block direction estimation ii) segmentation by direction intensity iii) Region of Interest (ROI) extraction by Morphological operations. Time Domain Method : To implement this method, we have considered four different features-namely ending, bifurcation, isolation and crossing points. To calculate the count of each feature in a fingerprint image, we have used 3*3 structuring elements which run through the entire 256*256 image. For comparison we use Euclidean distance model which calculates the Euclidean distance between feature vector of the test signature and feature vector of the database. The formula for Euclidean distance is given as follows: Let A (a1, a2 ... an) and B (b1, b2 bn) are two vectors of size n. We can calculate distance by equation 1 as:

Frequency Domain Method : For frequency domain method, we have considered DCT (Discrete Cosine Transform). The locations of four minutiae features in a particular fingerprint image are combined and represented by an image on which DCT is performed. DCT is a well-known signal analysis tool used in compression due to its compact representation power. Its known that Karhunen-Loeve transform (KLT) is the optimal transform in terms of information packing, however, its data dependent nature makes it infeasible to implement in some practical tasks. Moreover, DCT closely approximates the compact representation ability of the KLT, which makes it a very useful tool for signal representation both in terms of information packing

22

and in terms of computational complexity due to its data independent nature. DCT helps separate the image into parts (or spectral sub-bands) of differing importance (with respect to the image's visual quality). The general equation for a 1D (N data items) DCT is defined by the following equation:

*f(i)

Where, f (i) is the input sequence and

A(i) =
Minutiae extraction:-

for u = 1 and it is 0 otherwise

Minutiae extraction stage is divided in to two sub stages such as:- i) fingerprint ridge thinning and ii) minutiae marking. We used iterative parallel thinning algorithm for minutiae extraction stage. Ridge thinning is used to eliminate the redundant pixels of the ridges till the ridges are of one pixel wide. The minutiae marking is quite simple task. Here crossing number (CN) concept is used.

Post processing stage:For the post processing stage, it has only one sub step that is:- removal of false minutiae. Also a novel representation for bifurcations is proposed to unify terminations and bifurcation. The minutiae matcher determines whether the two minutiae sets are from the same finger or not. If the ridges are match well, then the two fingerprint images are aligned and matching is conducted for all remaining minutiae.

Fig 2.2.6 Images correctly detected as a match by the matching algorithm

23

2.2.2 Minutiae Matching:-

Fig 2.2.1 Minutiae Matcher

After testing the set of minutiae set of points of two finger print image, we perform Minutiae Matching to check whether they belong to the same person or not. It includes two consecutive stages: i) alignment stage ii) match stage i) Minutiae alignment:1) Let I1 & I2 be the two minutiae sets given by, I1= {m1,m2,..mM} where mi= (xi, yi, i) I2= {m1,m2,..mN} where mi= (xi, yi, i) The ridge associated with each minutiae is represented as a series of x-coordinates (x1, x2xn) of the points on the ridge. A point is sampled per ridge length L starting from the minutiae point, where the L is the average inter-ridge length. And n is set to 10 unless the total ridge length is less than 10*L. So the similarity of correlating the two ridges is derived from:

S=

At this stage (xi.xn) and (Xi.Xn) are the set of x-coordinates for the two minutiae which we have chosen. And the least possible of m is one of the value of n and N.

24

We will tally the score and if the score is greater than 0.8, then jump to step 2, if not then continue to match the next ridges pair. 2. Here we have to transform each set according to its own reference minutiae and then do match in a unified x-y coordinate. We are taking M ( x,y,) as reference minutiae which is in I1. We have to translate and rotate all other minutiae(xi, yi, i) for the finger prints we have taken into account with respect to the M (x,y,) according to the following formula: ( )=(TM) X [ ]

Where TM = [

The new coordinate system is originated at reference minutiae M and the new x-axis is coincident with the direction of minutiae M. Here scaling effect is not required, we are assuming two fingerprints taken from the same finger are having nearly the same size. Therefore we are getting transformed sets of minutiae I1' & I2'

Match stage:Generally the two identical minutiae are not exactly same due to the slight deformations and also inexact quantization. The algorithm for matching for the aligned minutiae patterns should be elastic. The minutiae matching elastic is done by keeping a bounding box around each of the template minutiae. If the minutiae which is to be matched is within that rectangle box and the direction discrepancy between them is so small, then the two minutiae are taken as a pair of matched minutiae. Each of the minutiae in that template image either has one corresponding minutiae or has no matched. The final match ratio for two fingerprints is given by
Match Score =

If the match score is greater than a threshold value which is pre-specified, then the two fingerprints taken are from the same finger.

25

Fig.2.2.6 One by one steps involved in fingerprint recognition algorithm

Thus the above explained diagram can be explained through an algorithm which is called as Algorithm of FRMSM : Input: Gray-scale Fingerprint image. Output: Verified fingerprint image with matching score.

26

1. Fingerprint is binarized 2. Thinning on binarized image 3. Minutiae points are extracted. Data matrix is generated to get the position, orientation and type of minutiae. 4. Matching of test fingerprint with template 5. Matching score of two images is computed, if matching score is 1 images are matched and if it is 0 then they are mismatched.

2.2.3Fingerprint sensors

SENSOR

MINUTIAE EXTRACTOR

MINUTIAE MATCHER

A fingerprint sensor is an electronic device used to capture a digital image of the fingerprint pattern. The captured image is called a live scan. This live scan is digitally processed to create a biometric template (a collection of extracted features) which is stored and used for matching. This is an overview of some of the more commonly used fingerprint sensor technologies.

Fig 2.3.1 Finger print image acquired by a Sensor

1.Optical Sensor

27

Optical fingerprint imaging involves capturing a digital image of the print using visible light. This type of sensor is, in essence, a specialized digital camera. The top layer of the sensor, where the finger is placed, is known as the touch surface. Beneath this layer is a light-emitting phosphor layer which illuminates the surface of the finger. The light reflected from the finger passes through the phosphor layer to an array of solid state pixels (a charge-coupled device) which captures a visual image of the fingerprint. A scratched or dirty touch surface can cause a bad image of the fingerprint. A disadvantage of this type of sensor is the fact that the imaging capabilities are affected by the quality of skin on the finger. For instance, a dirty or marked finger is difficult to image properly. Also, it is possible for an individual to erode the outer layer of skin on the fingertips to the point where the fingerprint is no longer visible. It can also be easily fooled by an image of a fingerprint if not coupled with a "live finger" detector. However, unlike capacitive sensors, this sensor technology is not susceptible to electrostatic discharge damage. Thus, Optical sensors are mainly used in fingerprint acquisition system. These sensors are of highly acceptable accuracy and high efficiency except for some cases like if the user's finger is too dirty or dry. 2.Ultrasonic Sensor Ultrasonic sensors make use of the principles of medical ultrasonography in order to create visual images of the fingerprint. Unlike optical imaging, ultrasonic sensors use very high frequency sound waves to penetrate the epidermal layer of skin. The sound waves are generated using piezoelectric transducers and reflected energy is also measured using piezoelectric materials. Since the dermal skin layer exhibits the same characteristic pattern of the fingerprint, the reflected wave measurements can be used to form an image of the fingerprint. This eliminates the need for clean, undamaged epidermal skin and a clean sensing surface. 3.Capacitance Sensor Capacitance sensors utilize the principles associated with capacitance in order to form fingerprint images. In this method of imaging, the sensor array pixels each act as one plate of a parallel-plate capacitor, the dermal layer (which is electrically conductive) acts as the other plate, and the non-conductive epidermal layer acts as a dielectric. 4.Passive capacitance Sensor A passive capacitance sensor uses the principle outlined above to form an image of the fingerprint patterns on the dermal layer of skin. Each sensor pixel is used to measure the capacitance at that point of the array. The capacitance varies between the ridges and valleys of the fingerprint due to the fact that the volume between the dermal layer and sensing element in valleys contains an air gap. The dielectric constant of the epidermis and the area of the sensing element are known values.

28

The measured capacitance values are then used to distinguish between fingerprint ridges and valleys. 5.Active capacitance Sensor Active capacitance sensors use a charging cycle to apply a voltage to the skin before measurement takes place. The application of voltage charges the effective capacitor. The electric field between the finger and sensor follows the pattern of the ridges in the dermal skin layer. On the discharge cycle, the voltage across the dermal layer and sensing element is compared against a reference voltage in order to calculate the capacitance. The distance values are then calculated mathematically, and used to form an image of the fingerprint. Active capacitance sensors measure the ridge patterns of the dermal layer like the ultrasonic method. Again, this eliminates the need for clean, undamaged epidermal skin and a clean sensing surface.

29

Fig. 11 3X3 neighbourhood

30

4.3 Fingerprint Databases

Research in biometrics profoundly depends on the availability of sensed data. The growth that the field has experienced over the past two decades has led to the appearance of increasing numbers of biometric databases, either monomodal (one biometric trait sensed) or multimodal (two or more biometric traits sensed). Previous to the International Fingerprint Verification Competitions (FVC), the only large, publicly available datasets were the NIST databases. However, thesedatabases were not well suited for the evaluation of algorithms operating with livescan images and will not be described here. In this section, we present some of the most important publicly available biometric databases, either monomodal or multimodal, that include the fingerprint trait acquired with live-scan sensors.Some of the important databases are1. Fingerprint Verification Competitions (FVC) database 2. 3. 4. 5.

MCYT Bimodal Database BIOMET Multimodal Database Michigan State University (MSU) Database BiosecurID Multimodal Database

behavioral biometric parameters are signature, gait, speech and keystroke, these parameters change with age and environment. However physiological characteristics such as face, fingerprint, palm print and iris remains unchanged through out the life time of a person. The biometric system operates as verification mode or identification mode depending on the requirement of an application. The verification mode validates a persons identity by comparing captured biometric data with ready made template. The identification mode recognizes a persons identity by performing matches against multiple fingerprint biometric templates.

1.9 The Road Ahead

Market estimates put the total hardware sales for the industry at US$6.6 million in 1990 and nearly US$200 million in 2000.Whether the next decade will result in a similar 2500% increase will depend upon user demand for positive identification biometrics.That demand will be created by imaginatively created systems designed for convenience, friendliness, cost-effectiveness and ease of use. The use of negative identification biometrics will be fueled by government requirements to limit citizens to a single identity in driver licensing,

31

social service and other civil applications [67, 68]. That demand will require the development of stronger criteria for cost/benefit assessment, security assurance,and privacy protection.Although we cannot predict the future rate of growth of the industry with any certainty, we do know that long-term growth is inevitable. With this book, we hope to stimulate further inquiry into the technologies, applications and issues that will shape this industry in the years to come.
The approaches are distinguished on the basis of several factors like: the kind of input images they handle i.e. whether binary or gray scale, techniques of binarization and segmentation involved, whether thinning is required or not and the amount of effort required in the post processing stage, if exists. But low quality fingerprint images need preprocessing to increase contrast, and reduce different types of noises as noisy pixels also generate a lot of spurious minutiae as they also get enhanced during the pre processing steps. Further, more emphasis is to be laid on defining the local criteria, in order to establish the validity of a minutiae point, which is particularly useful during fingerprint matching and adopting more sophisticated identification models, for instance extending minutiae definition by including trifurcations, islands, bridges, spurs etc. Also, the paper leads to the further study of the statistical theory of fingerprint minutiae. In particular approaches can be investigated to determine the number of degrees of freedom within a fingerprint population which will give a sound understanding of the statistical uniqueness of fingerprint minutiae.

Performance evaluation index:There are two types performance evaluation indexes to determine the performance of a fingerprint recognition system such as:7.1.1 False Rejection Rate (FRR): Sometimes the biometric security system may incorrectly reject an access attempt by an authorized user. To measure these types of incidents FAR is basically used. A systems FRR basically states the ratio between the number of false rejections and the number of identification attempts.

FRR (%) FRR= (FR/N)*100 FR=number of incidents of false rejections N= number of samples

5. Conclusions Image quality is related directly to the ultimate performance of automatic fingerprint authentication systems. Good quality fingerprint images need only minor pre processing and enhancement for accurate feature detection algorithm.

32