Attendance System

Student attention remains ground On Fingerprint Recognition and iodin-to-M some(prenominal) coordinated A dissertation submitted in partial ful? llment of the requirements for the form of Bachelor of computing machine Application in Computer wisdom by Sachin ( debate no. 107cs016) and Arun Sharma (Roll no. 107cs015) Under the guidance of Prof. R. C. Tripathi department of Computer Science and applied science National impart of Technology Rourkela Rourkela-769 008, Orissa, India 2 . Dedicated to Our P bents and Indian Scienti? c Community . 3 National Institute of Technology Rourkela Certi? cateThis is to certify that the project entitled, Student attention System establish On Fingerprint Recognition and One-to- m distributively an otherwise(prenominal) Matching submitted by Rishabh Mishra and Prashant Trivedi is an au and sotic cast carried out by them under my supervision and guidance for the partial ful? llment of the requirements for the pillage of Bachelor of Tec hnology Degree in Computer Science and plan at National Institute of Technology, Rourkela. To the best of my knowledge, the matter embodied in the project has non been submitted to any other(a) University / Institute for the award of any Degree or Diploma.Date 9/5/2011 Rourkela (Prof. B. Majhi) Dept. of Computer Science and Engineering 4 Abstract Our project aims at designing an student attending dodge which could e? ectively manage attention of students at workings like NIT Rourkela. Attendance is marked after student identi? cation. For student identi? cation, a ? ngerprint recognition base identi? cation dodge is usaged. Fingerprints be considered to be the best and fastest method for biometric identi? cation. They argon secure to function, unique for e rattling person and does not change in ones life while. Fingerprint recognition is a mature ? ld straightaway, just now s gutter identifying individual from a particularise of enrolled ? ngerprints is a quantif y pickings dish. It was our responsibility to mitigate the ? ngerprint identi? cation system for implementation on heroic databases e. g. of an institute or a country etc. In this project, legion(predicate) advanced algorithms confuse been utilise e. g. gender regard, primaeval establish one to many inter affiliated, removing bourn minutiae. apply these new algorithms, we have real an identi? cation system which is smart in implementation than any other functional today in the market. Although we argon con labor unionption this ? ngerprint identi? cation system for student identi? ation purpose in our project, the co-ordinated results atomic morsel 18 so good that it could perform very well on greathearted databases like that of a country like India (MNIC jut out). This system was implemented in Matlab10, Intel Core2Duo make foror and comparison of our one to many identi? cation was do with animated identi? cation proficiency i. e. one to one identi? cation on uniform platform. Our check outing technique runs in O(n+N) season as comp ard to the existing O(Nn2 ). The ? ngerprint identi? cation system was tested on FVC2004 and Veri? nger databases. 5 Acknowledgments We present our profound gratitude and indebtedness to Prof. B.Majhi, Department of Computer Science and Engineering, NIT, Rourkela for introducing the evince topic and for their inspiring intellectual guidance, constructive criticism and priceless suggestion throughout the project work. We ar also thankful to Prof. Pankaj Kumar Sa , Ms. Hunny Mehrotra and other sta? s in Department of Computer Science and Engineering for need us in up the algorithms. Fin tot every last(predicate)yy we would like to thank our p bents for their substantiate and permitting us stay for more days to complete this project. Date 9/5/2011 Rourkela Rishabh Mishra Prashant Trivedi Contents 1 Introduction 1. 1 1. 2 1. 3 1. 4 1. 1. 6 1. 7 problem Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . Motivation and Challenges . . . . . . . . . . . . . . . . . . . . . . . . victimisation Biometrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . What is ? ngerprint? . . . . . . . . . . . . . . . . . . . . . . . . . . . why use ? ngerprints? . . . . . . . . . . . . . . . . . . . . . . . . . . . exploitation ? ngerprint recognition system for attention counsel . . . Organization of the thesis . . . . . . . . . . . . . . . . . . . . . . . . 17 17 17 18 18 19 19 19 21 21 22 23 24 24 30 30 33 33 33 35 35 36 36 2 Attendance Management Framework 2. 2. 2 2. 3 2. 4 2. 5 picturer ironw atomic bod 18 softwargon program take aim Design . . . . . . . . . . . . . . . . . . . . Attendance Management Approach . . . . . . . . . . . . . . . . . . . on-line(a) Attendance topic Generation . . . . . . . . . . . . . . . . . Ne some(prenominal)rk and selective selective studybase Management . . . . . . . . . . . . . . . . . . utilise radio interlocking instead of local atomic number 18a nedeucerk and bringing portability . . . 2. 5. 1 2. 6 U evilg Por get across ruse . . . . . . . . . . . . . . . . . . . . . . Comparison with other student attending systems . . . . . . . . . . 3 Fingerprint Identi? cation System 3. 1 3. 2 How Fingerprint Recognition whole kit and boodle? . . . . . . . . . . . . . . . . . Fingerprint Identi? cation System Flowchart . . . . . . . . . . . . . . 4 Fingerprint Enhancement 4. 1 4. 2 4. 3 naval division . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . standardization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . predilection estimation . . . . . . . . . . . . . . . . . . . . . . . . . . 6 CONTENTS 4. 4 4. 5 4. 6 4. 7 c everyplaceline relative absolute frequency estimation . . . . . . . . . . . . . . . . . . . . . . . Gabor ? lter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Binarisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Thinning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 38 39 40 40 41 41 42 42 43 44 45 45 45 46 47 47 50 51 53 53 54 54 55 56 57 59 59 59 59 60 5 maneuver characteristic rootage 5. 1 5. 2 decision the Reference beat period . . . . . . . . . . . . . . . . . . . . . . . Minutiae fall and Post-Processing . . . . . . . . . . . . . . . . 5. 2. 1 5. 2. 2 5. 2. 3 5. 3 Minutiae lineage . . . . . . . . . . . . . . . . . . . . . . . Post-Processing . . . . . . . . . . . . . . . . . . . . . . . . . Removing Boundary Minutiae . . . . . . . . . . . . . . . . . . Extraction of the report . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. 3. 1 What is paint? . . . . . . . . . . . . . . . . . . . . . . . . . . Simple differentiate . . . . . . . . . . . . . . . . . . . . . . . . . . . . Complex find . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 variance of selective informationbase 6. 1 6. 2 6. 3 sex adherence . . . . . . . . . . . . . . . . . . . . . . . . . . . . consorti? cation of Fingerprint . . . . . . . . . . . . . . . . . . . . . . . naval division . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Matching 7. 1 7. 2 7. 3 Alignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . alive Matching Techniques . . . . . . . . . . . . . . . . . . . . . One to umteen twin(a) . . . . . . . . . . . . . . . . . . . . . . . . . . 7. 3. 1 7. 4 7. 5 order of One to Many Matching . . . . . . . . . . . . . . . do distinguish match and broad interconnected . . . . . . . . . . . . . . . . Time Complexity of this co-ordinated technique . . . . . . . . . . . . . . 8 Experimental depth psychology 8. 1 8. 2 murder Environment . . . . . . . . . . . . . . . . . . . . . . Fingerprint Enhancement . . . . . . . . . . . . . . . . . . . . . . . . 8. 2. 1 8. 2. 2 segmentation and Normalization . . . . . . . . . . . . . . . . Orientation Estimation . . . . . . . . . . . . . . . . . . . . . . 8 8. 2. 3 8. 2. 4 8. . 5 8. 3 CONTENTS rooftree relative frequency Estimation . . . . . . . . . . . . . . . . . . . Gabor Filters . . . . . . . . . . . . . . . . . . . . . . . . . . . Binarisation and Thinning . . . . . . . . . . . . . . . . . . . . 60 60 61 62 62 62 63 64 64 64 64 65 66 66 Feature Extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. 3. 1 Minutiae Extraction and Post Processing . . . . . . . . . . . . Minutiae Extraction . . . . . . . . . . . . . . . . . . . . . . . later on Removing inau whencetic and Boundary Minutiae . . . . . . . 8. 3. 2 Reference Point Detection . . . . . . . . . . . . . . . . . . . . 8. 4 grammatical gender Estimation and Classi? ation . . . . . . . . . . . . . . . . . . 8. 4. 1 8. 4. 2 sex activity Estimation . . . . . . . . . . . . . . . . . . . . . . . . Classi? cation . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. 5 8. 6 Enrolling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Matching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. 6. 1 8. 6. 2 Fingerprint Veri? cation extends . . . . . . . . . . . . . . . . . Identi? cation Results and Comparison with Other Matching techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 70 73 74 75 75 79 8. 7 work Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Conclusion 9. 1 Outcomes of this Project . . . . . . . . . . . . . . . . . . . . . . . . . 10 Future calculate and Expectations 10. 1 Approach for Future Work A Matlab functions . . . . . . . . . . . . . . . . . . . . . . . List of human bodys 1. 1 2. 1 2. 2 2. 3 2. 4 2. 5 2. 6 2. 7 2. 8 3. 1 4. 1 4. 2 Example of a continue ending and a bifurcation . . . . . . . . . . . . . . Hardwargon present in elucidaterooms . . . . . . . . . . . . . . . . . . . . . Classroom Scenario . . . . . . . . . . . . . . . . . . . . . . . . . . . . profit Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ER Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Level 0 DFD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Level 1 DFD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Level 2 DFD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Por turn off tress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fingerprint Identi? cation System Flowchart . . . . . . . . . . . . . . Orientation Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . (a)Original two-baser, (b) heighten theatrical role, (c)Binarised run into, (d)Thinned Image . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. 1 Row 1 ? lter retort c1k , k = 3, 2, and 1. Row 2 ? lter rejoinder c2k , k = 3, 2, and 1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. 2 5. 3 Examples of (a) continue-ending (CN=1) and (b)bifurcation pel (CN=3) 42 43 40 18 22 23 25 26 27 27 28 29 34 37 Examples of typical morose minutiae structures (a)Spur, (b)Hole, (c)Triangle, (d)Spike . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 44 44 45 48 5. 4 5. 5 5. 6 6. 1 Skeleton of windowpane centered at margin minutiae . . . . . . . . . . hyaloplasm Representation of boundary minutiae . . . . . . . . . . . . . cay Representation . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gender Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 10 6. 2 6. 3 LIST OF FIGURES 135o occlusions of a ? ngerprint . . . . . . . . . . . . . . . . . . . . . . . . Fingerprint Classes (a)Left Loop, (b)Right Loop, (c)Whorl, (d1) bend, (d2)Tented Arch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. 4 7. 1 8. 1 8. 2 8. 3 8. 4 8. 5 8. 6 8. 7 8. 8 8. 9 divide Database . . . . . . . . . . . . . . . . . . . . . . . . . . One to Many Matching . . . . . . . . . . . . . . . . . . . . . . . . . . Normalized Image . . . . . . . . . . . . . . . . . . . . . . . . . . . . Orientation Image . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ridge Frequency Image . . . . . . . . . . . . . . . . . . . . . . . . . . Left-Original Image, Right- raise Image . . . . . . . . . . . . . . Binarised Image . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Thinned Image . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . entirely Extracted Minutiae . . . . . . . . . . . . . . . . . . . . . . . . . . manifold Image with spurious and boundary minutiae . . . . . . . . Minutiae Image after post-processing . . . . . . . . . . . . . . . . . 51 52 57 59 60 60 61 61 62 62 63 63 64 65 50 8. 10 Composite Image after post-processing . . . . . . . . . . . . . . . . . 8. 11 Plotted Minutiae with Reference Point(Black Spot) . . . . . . . . . . 8. 12 represent Time interpreted for Identi? cation vs surface of Database(key based one to many identi? cation) . . . . . . . . . . . . . . . . . . . . . . . . 8. 13 interpret Time taken for Identi? cation vs Size of Database (n2 ide nti? cation) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. 14 Expected represent for comparison Time taken for Identi? cation vs Size of Database(1 million) . . . . . . . . . . . . . . . . . . . . . . . . . 68 69 71 List of Tables 2. 1 5. 1 8. 1 8. 2 8. 3 8. 4 8. 5 8. 6 8. 7 8. 8 Estimated calculate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Properties of Crossing physical body . . . . . . . . . . . . . . . . . . . . . 22 43 64 65 66 66 67 67 68 fair(a) tot of Minutiae before and after post-processing . . . . Ridge assiduity Calculation Results . . . . . . . . . . . . . . . . . . . . Classi? cation Results on Original Image . . . . . . . . . . . . . . . . Classi? cation Results on Enhanced Image . . . . . . . . . . . . . . . Time taken for Classi? cation . . . . . . . . . . . . . . . . . . . . . . .Time taken for Enrolling . . . . . . . . . . . . . . . . . . . . . . . . . Error Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Performance of ours and n2 matching based identi? cation techniques on a database of size 150 . . . . . . . . . . . . . . . . . . . . . . . . . 70 11 List of Algorithms 1 2 3 4 Key Extraction Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . Gender Estimation Algorithm . . . . . . . . . . . . . . . . . . . . . . . Key ground One to Many Matching Algorithm . . . . . . . . . . . . . . Matching Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 49 55 56 12Chapter 1 Introduction 1. 1 Problem Statement Designing a student attending management system based on ? ngerprint recognition and faster one to many identi? cation that manages records for attending in institutes like NIT Rourkela. 1. 2 Motivation and Challenges Every organization whether it be an educational institution or business organization, it has to maintain a becoming record of attention of students or employees for e? ective functioning of organization. Designing a better attend ing management system for students so that records be retained with ease and accuracy was an important key behind motivating this project.This would improve accuracy of attention records because it go away despatch all the hassles of roll life history and depart save valuable time of the students as well as t all(prenominal)ers. Image processing and ? ngerprint recognition be very advanced today in terms of technology. It was our responsibility to improve ? ngerprint identi? cation system. We decreased matching time by partitioning the database to one-tenth and improved matching apply key based one to many matching. 13 14 CHAPTER 1. INTRODUCTION 1. 3 exploitation Biometrics Biometric Identi? cation Systems ar astray apply for unique identi? cation of valet de chambre mainly for veri? cation and identi? ation. Biometrics is apply as a form of identity operator access management and access control. So use of biometrics in student attendance management system is a secur e approach. thither be many typewrites of biometric systems like ? ngerprint recognition, face recognition, voice recognition, iris recognition, handle recognition etc. In this project, we apply ? ngerprint recognition system. 1. 4 What is ? ngerprint? A ? ngerprint is the pattern of ridges and valleys on the surface of a ? ngertip. The end sharpens and crossing points of ridges are called minutiae. It is a widely take awayed assumption that the minutiae pattern of each ? ger is unique and does not change during ones life. Ridge endings are the points where the ridge tailor terminates, and bifurcations are where a ridge splits from a single path to two paths at a Y-junction. project 1 illustrates an example of a ridge ending and a bifurcation. In this example, the black pels stop to the ridges, and the clear pixels correspond to the valleys. embodiment 1. 1 Example of a ridge ending and a bifurcation When human ? ngerprint experts determine if two ? ngerprints are from th e same ? nger, the matching degree in the midst of two minutiae pattern is one of the close to important factors.Thanks to the affinity to the way of human ? ngerprint experts and compactness of templates, the minutiae-based matching method is the most widely studied matching method. 1. 5. WHY USE fingermarkS? 15 1. 5 Why use ? ngerprints? Fingerprints are considered to be the best and fastest method for biometric identi? cation. They are secure to use, unique for either person and does not change in ones lifetime. Besides these, implementation of ? ngerprint recognition system is cheap, prosperous and accurate up to satis? ability. Fingerprint recognition has been widely used in twain forensic and civilian applications.Compared with other biometrics gives , ? ngerprint-based biometrics is the most proven technique and has the largest market shares . Not only it is faster than other techniques alone also the energy consumption by such systems is every(prenominal)placely l ess. 1. 6 Using ? ngerprint recognition system for attendance management Managing attendance records of students of an institute is a tedious task. It consumes time and paper both. To make all the attendance tie in work automatic and on-line, we have designed an attendance management system which could be implemented in NIT Rourkela.It uses a ? ngerprint identi? cation system developed in this project. This ? ngerprint identi? cation system uses existing as well as new techniques in ? ngerprint recognition and matching. A new one to many matching algorithm for large databases has been introduced in this identi? cation system. 1. 7 Organization of the thesis This thesis has been organized into ten chapters. Chapter 1 introduces with our project. Chapter 2 explains the proposed design of attendance management system. Chapter 3 explains the ? ngerprint identi? cation system used in this project.Chapter 4 explains sweetener techniques, Chapter 5 explains feature filiation methods, Ch apter 6 explains our database partitioning approach . Chapter 7 explains matching technique. Chapter 8 explains experimental work done and performance analysis. Chapter 9 includes conclusions and Chapter 10 introduces proposed future work. Chapter 2 Attendance Management Framework manual attendance taking and report generation has its limitations. It is well enough for 30-60 students but when it comes to taking attendance of students large in number, it is di? cult. For taking attendance for a lecture, a conference, etc. oll calling and manual attendance system is a failure. Time excess over answers of students, waste of paper etc. are the disadvantages of manual attendance system. Moreover, the attendance report is also not generated on time. Attendance report which is circulated over NITR webmail is two months old. To overcome these non-optimal situations, it is essential that we should use an automatic on-line attendance management system. So we present an implemen sidestep a ttendance management framework. Student attendance system framework is divided into tether parts Hardware/Software Design, Attendance Management Approach and on-line(a) Report Generation. distributively of these is explained below. 2. 1 Hardware Software Level Design take hardware used should be easy to maintain, implement and easily available. Proposed hardware consists succeeding(a) parts (1)Fingerprint S brush asidener, (2)LCD/Display Module (optional), (3)Computer 16 2. 2. attendance anxiety APPROACH Table 2. 1 Estimated Budget Device Cost of Number of total Name One Unit Units Required Unit Budget S targetner 500 100 50000 PC 21000 100 2100000 Total 21,50,000 (4)local area network data link 17 Fingerprint s rear endner leave be used to excitant ? ngerprint of instructors/students into the computer software.LCD present lead be displaying rolls of those whose attendance is marked. Computer Software go away be interfacing ? ngerprint s bunsner and LCD and depart be connected to the network. It lead arousal ? ngerprint, will process it and extract features for matching. after(prenominal) matching, it will update database attendance records of the students. project 2. 1 Hardware present in varietyrooms Estimated Budget Estimated cost of the hardware for implementation of this system is shown in the table 2. 1. Total number of classrooms in NIT Rourkela is more or less 100. So number of units necessary will be 100. 2. 2 Attendance Management ApproachThis part explains how students and teachers will use this attendance management system. Following points will make legitimate that attendance is marked correctly, without any problem (1)All the hardware will be interior classroom. So outside interference will be absent. (2)To annihilate unauthorized access and unwanted attempt to corrupt the hardware by students, all the hardware except ? ngerprint s usher outner could be put inside a small 18 CHAPTER 2. ATTENDANCE MANAGEMENT material cab in. As an alternate solution, we give the gate install CCTV cameras to hold back unprivileged activities. (3)When teacher enters the classroom, the attendance scrape will start.Computer software will start the process after arousalting ? ngerprint of teacher. It will ? nd the Subject ID, and Current Semester exploitation the ID of the teacher or could be right manually on the software. If teacher doesnt enter classroom, attendance marking will not start. (4)After some time, say 20 proceedings of this process, no attendance will be exitn because of late entrance. This time period can be increased or decreased as per requirements. get in 2. 2 Classroom Scenario 2. 3 On-Line Attendance Report Generation Database for attendance would be a table having pursual ? elds as a combination for primary ? ld (1)Day,(2)Roll,(3)Subject and following non-primary ? elds (1)Attendance,(2)Semester. Using this table, all the attendance can be managed for a student. For on-line report generati on, a unprejudiced website can be hosted on NIT Rourkela servers, 2. 4. communicate AND DATABASE MANAGEMENT 19 which will access this table for showing attendance of students. The sql queries will be used for report generation. Following inquiry will give total amount of classes held in subject CS423 SELECT COUNT(DISTINCT Day) FROM AttendanceTable WHERE SUBJECT = CS423 AND Attendance = 1 For attendance of oll 107CS016, against this subject, following call into question will be used SELECT COUNT(Day) FROM AttendanceTable WHERE Roll = 107CS016 AND SUBJECT = CS423 AND Attendance = 1 Now the attendance percent can easily be calculated ClassesAttended ? 100 ClassesHeld Attendance = (2. 1) 2. 4 Network and Database Management This attendance system will be spread over a wide network from classrooms via intranet to net income. Network diagram is shown in ? g. 2. 3. Using this network, attendance reports will be made available over internet and e-mail. A monthly report will be sent t o each student via email and website will show the updated attendance.Entity relationship diagram for database of students and attendance records is shown in ? g. 2. 4. In ER diagram, primary ? elds are Roll, Date, SubjectID and TeacherID. Four tables are Student, Attendance, Subject and Teacher. Using this database, attendance could easily be maintained for students. Data? ow is shown in data ? ow diagrams (DFD) shown in ? gures 2. 5, 2. 6 and 2. 7. 2. 5 Using receiving rigid network instead of LAN and bringing portability We are using LAN for communication among servers and hardwares in the classrooms. We can instead use radiocommunication LAN with portable inventions.Portable braid will have an embedded ? ngerprint scanner, wireless connection, a microprocessor loaded with a software, recollection and a display terminal, see ? gure 2. 5. Size of widget could be small like a erratic speech sound depending upon how well the gimmick is manufactured. 20 CHAPTER 2. ATTENDANCE MANAGEMENT FRAMEWORK Figure 2. 3 Network Diagram 2. 5. USING WIRELESS NETWORK quite OF LAN AND BRINGING PORTABILITY21 Figure 2. 4 ER Diagram 22 CHAPTER 2. ATTENDANCE MANAGEMENT FRAMEWORK Figure 2. 5 Level 0 DFD Figure 2. 6 Level 1 DFD 2. 5. USING WIRELESS NETWORK INSTEAD OF LAN AND BRINGING PORTABILITY23 Figure 2. Level 2 DFD 24 CHAPTER 2. ATTENDANCE MANAGEMENT FRAMEWORK This device should have a wireless connection. Using this wireless connection, Figure 2. 8 Portable Device attendance taken would be updated automatically when device is in network of the nodes which are storing the attendance records. Database of enrolled ? ngerprints will be in this portable device. Size of enrolled database was 12. 1 MB when 150 ? ngerprints were enrolled in this project. So for 10000 students, atleast 807 MB or more space would be required for storing enrolled database. For this purpose, a removable memory chip could be used.We cannot use wireless LAN here because transport data using wirele ss LAN will not be possible because of less range of wireless devices. So enrolled data would be on chip itself. Attendance results will be updated when portable device will be in the range of nodes which are storing attendance reports. We may update all the records online via the mobile network provided by di? erent companies. Today 3G network provides su? cient throughput which can be used for updating attendance records automatically without firing near nodes. In such case, 2. 6. COMPARISON WITH OTHER STUDENT ATTENDANCE SYSTEMS 25 he need of database inside memory chip will not be mandatory. It will be fetched by using 3G mobile network from central database repository. The design of such a portable device is the task of embedded system engineers. 2. 5. 1 Using Portable Device In this section, we suggest the working of portable device and the method of using it for marking attendance. The device may either be having touchscreen input/display or buttons with lcd display. A softw are specially designed for the device will be running on it. Teachers will verify his/her ? ngerprint on the device before giving it to students for marking attendance.After verifying the teachers identity, software will ask for course and and other required information about the class which he or she is going to teach. Software will ask teacher the time after which device will not mark any attendance. This time can vary depending on the teachers mood but our suggested esteem is 25 minutes. This is done to prevent late entrance of students. This mistreat will hardly take a couple of(prenominal) seconds. and then students will be given device for their ? ngerprint identi? cation and attendance marking. In the continuation, teacher will start his/her lecture.Students will hand over the device to other students whose attendance is not marked. After 25 minutes or the time decided by teacher, device will not input any attendance. After the class is over, teacher will take device and will end the lecture. The main function of software running on the device will be ? ngerprint identi? cation of students followed by report generation and direct reports to servers using 3G network. Other functions will be downloading and updating the database available on the device from central database repository. 2. 6 Comparison with other student attendance systemsThere are various other kind of student attendance management systems available like RFID based student attendance system and GSM-GPRS based student attendance system. These systems have their own pros and cons. Our system is better because ? rst it saves time that could be used for teaching. Second is portability. Portability 26 CHAPTER 2. ATTENDANCE MANAGEMENT FRAMEWORK has its own advantage because the device could be taken to any class wherever it is scheduled. While GSM-GPRS based systems use fix of class for attendance marking which is not dynamic and if schedule or location of the class changes, wrong attenda nce might be marked.Problem with RFID based systems is that students have to reach out RFID cards and also the RFID detectors are needed to be installed. Nonetheless, students may give proxies easily using friends RFID card. These problems are not in our system. We used ? ngerprints as recognition criteria so proxies cannot be given. If portable devices are used, attendance marking will be done at any place and any time. So our student attendance system is faraway better to be implemented at NITR. Chapter 3 Fingerprint Identi? cation System An identi? cation system is one which helps in identifying an individual among any people when precise information is not available. It may involve matching available features of prognosis like ? ngerprints with those already enrolled in database. 3. 1 How Fingerprint Recognition works? Fingerprint simulacrums that are found or scanned are not of optimal quality. So we remove echos and enhance their quality. We extract features like minuti ae and others for matching. If the sets of minutiae are matched with those in the database, we call it an identi? ed ? ngerprint. After matching, we perform post-matching go which may include showing details of identi? ed candidate, marking attendance etc.A drawing ? owchart is shown in next section. 3. 2 Fingerprint Identi? cation System Flowchart A brief methodology of our Fingerprint Identi? cation System is shown here in following ? owchart. Each of these are explained in the later chapters. 27 28 CHAPTER 3. FINGERPRINT acknowledgement SYSTEM Figure 3. 1 Fingerprint Identi? cation System Flowchart Chapter 4 Fingerprint Enhancement The mental word-painting acquired from scanner is sometimes not of finished quality . It gets corrupted collectible to irregularities and non-uniformity in the impression taken and due to variations in the skin and the presence of the scars, humidity, irt etc. To overcome these problems , to reduce noise and enhance the de? nition of ridges agai nst valleys, various techniques are applied as following. 4. 1 Segmentation Image segmentation 1 disperses the foreground regions and the home regions in the image. The foreground regions refers to the take ? ngerprint area which ends the ridges and valleys. This is the area of interest. The background regions refers to the regions which is outside the borders of the main ? ngerprint area, which does not contain any important or valid ? ngerprint information.The extraction of noisy and unreasonable minutiae can be done by applying minutiae extraction algorithm to the background regions of the image. Thus, segmentation is a process by which we can discard these background regions, which results in more reliable extraction of minutiae points. We are going to use a method based on variance thresholding . The background regions usher a very low grey scale variance jimmy , whereas the foreground regions have a very high variance . for the first time , the image is divided into seal offs and the grey-scale variance is calculated for each pig out in the image .If the variance is less than the global threshold , then the scarf out is assigned to be part of background region or else 29 30 CHAPTER 4. FINGERPRINT sweetening it is part of foreground . The grey direct variance for a third power of size S x S can be calculated as 1 V ar(k) = 2 S S? 1 S? 1 (G(i, j) ? M (k))2 i=0 j=0 (4. 1) where Var(k) is the grey take aim variance for the block k , G(i,j) is the grey level value at pixel (i,j) , and M(k) denotes the think of grey level value for the corresponding block k . 4. 2 Normalization Image normalization is the next step in ? ngerprint sweetener process.Normalization 1 is a process of standardizing the intensity value in an image so that these intensity values lies indoors a certain desired range. It can be done by adjusting the range of grey-level values in the image. permit G(i, j) denotes the grey-level value at pixel (i, j), and N(i, j) represent the normalized grey-level value at pixel (i, j). Then the normalized image can de? ned as ? ? M + 0 N (i, j) = ? M ? 0 V0 (G(i,j)? M )2 V V0 (G(i,j)? M )2 V , if I(i, j) > M , otherwise where M0 and V0 are the estimated mean and variance of I(i, j), respectively . 4. 3 Orientation estimation The orientation course ? eld of a ? ngerprint image de? es the local orientation of the ridges contained in the ? ngerprint . The orientation estimation is a fundamental step in the enhancement process as the subsequent Gabor ? ltering stage relies on the local orientation in order to e? ectively enhance the ? ngerprint image. The least mean square estimation method used by Raymond Siamese 1 is used to compute the orientation image. However, instead of estimating the orientation block-wise, we have elect to extend their method into a pixel-wise scheme, which set ups a ? ner and more accurate estimation of the orientation ? eld. The steps for calculating the orientation at pixel i, j) are as follows 4. 3. ORIENTATION ESTIMATION 31 1. Firstly , a block of size W x W is centered at pixel (i, j) in the normalized ? ngerprint image. 2. For each pixel in the block, compute the gradients dx (i, j) and dy (i, j), which are the gradient magnitudes in the x and y directions, respectively. The horizontal Sobel operator6 is used to compute dx(i, j) 1 0 -1 2 0 -21 0 -1 Figure 4. 1 Orientation Estimation 3. The local orientation at pixel (i j) can then be estimated using the following equations i+ W 2 j+ W 2 Vx (i, j) = u=i? W 2 i+ W 2 v=j? W 2 j+ W 2 2? x (u, v)? y (u, v) (4. 2) Vy (i, j) = u=i? W v=j? W 2 2 2 2 ? (u, v) ? ?y (u, v), (4. 3) ?(i, j) = 1 Vy (i, j) tan? 1 , 2 Vx (i, j) (4. 4) where ? (i, j) is the least square estimate of the local orientation at the block centered at pixel (i, j). 4. Smooth the orientation ? eld in a local neighborhood using a Gaussian ? lter. The orientation image is ? rstly converted into a continuous vector ? eld, which is de? ned as ? x (i, j) = cos 2? (i, j), ? y (i, j) = sin 2? (i, j), (4. 5) (4. 6) where ? x and ? y are the x and y components of the vector ? eld, respectively. After 32 CHAPTER 4. FINGERPRINT ENHANCEMENT the vector ? eld has been computed, Gaussian smoothing is then performed as follows w? w? 2 ?x (i, j) = w? u=? 2 w? v=? 2 G(u, v)? x (i ? uw, j ? vw), (4. 7) w? 2 w? 2 ?y (i, j) = w? u=? 2 w? v=? 2 G(u, v)? y (i ? uw, j ? vw), (4. 8) where G is a Gaussian low-pass ? lter of size w? x w? . 5. The ? nal smoothed orientation ? eld O at pixel (i, j) is de? ned as O(i, j) = ? y (i, j) 1 tan? 1 2 ? x (i, j) (4. 9) 4. 4 Ridge Frequency Estimation other important parameter,in addition to the orientation image, that can be used in the construction of the Gabor ? lter is the local ridge frequency. The local frequency of the ridges in a ? ngerprint is represented by the frequency image. The ? st step is to divide the image into blocks of size W x W. In the next step we project the greylevel values of ea ch pixels located inside each block on a direction perpendicular to the local ridge orientation. This projection results in an virtually sinusoidal-shape wave with the local minimum points denoting the ridges in the ? ngerprint. It involves smoothing the projected waveform using a Gaussian lowpass ? lter of size W x W which helps in cut down the e? ect of noise in the projection. The ridge space S(i, j) is then calculated by counting the median number of pixels between the consecutive minima points in the projected waveform.The ridge frequency F(i, j) for a block centered at pixel (i, j) is de? ned as F (i, j) = 1 S(i, j) (4. 10) 4. 5. GABOR extend 33 4. 5 Gabor ? lter Gabor ? lters 1 are used because they have orientation-selective and frequencyselective properties. Gabor ? lters are called the induce of all other ? lters as other ? lter can be derived using this ? lter. Therefore, applying a properly tuned Gabor ? lter can preserve the ridge structures while bring down nois e. An even-symmetric Gabor ? lter in the spatial domain is de? ned as 1 x2 y2 G(x, y, ? , f ) = exp? ? + ? cos 2? f x? , 2 2 2 ? x ? y (4. 11) x? = x cos ? + y sin ? , (4. 12) y? ? x sin ? + y cos ? , (4. 13) where ? is the orientation of the Gabor ? lter, f is the frequency of the cosine wave, ? x and ? y are the standard deviations of the Gaussian gasbag along the x and y axes, respectively, and x? and y? de? ne the x and y axes of the ? lter coordinate frame respectively. The Gabor Filter is applied to the ? ngerprint image by spatially convolving the image with the ? lter. The convolution of a pixel (i,j) in the image requires the corresponding orientation value O(i,j) and the ridge frequency value F(i,j) of that pixel . wy 2 wx 2 E(i, j) = u=? wx 2 w v=? 2y G(u, v, O(i, j), F (i, j))N (i ? u, j ? v), (4. 4) where O is the orientation image, F is the ridge frequency image, N is the normalized ? ngerprint image, and wx and wy are the width and height of the Gabor ? lter mask, respectively. 34 CHAPTER 4. FINGERPRINT ENHANCEMENT 4. 6 Binarisation Most minutiae extraction algorithms operate on basically binary images where on that point are only two levels of interest the black pixels represent ridges, and the white pixels represent valleys. Binarisation 1 converts a greylevel image into a binary image. This helps in improving the contrast between the ridges and valleys in a ? ngerprint image, and consequently facilitates the extraction of minutiae.One very useful property of the Gabor ? lter is that it contains a DC component of zero, which indicates that the resulting ? ltered image has a zero mean pixel value. Hence, binarisation of the image can be done by using a global threshold of zero. Binarisation involves examining the grey-level value of every pixel in the deepen image, and, if the grey-level value is greater than the prede? ned global threshold, then the pixel value is set to value one else, it is set to zero. The burden of binarisation is a binary image which contains two levels of information, the background valleys and the foreground ridges. . 7 Thinning Thinning is a morphological carrying out which is used to remove selected foreground pixels from the binary images. A standard thinning algorithm from 1 is used, which performs this operation using two subiterations. The algorithm can be accessed by a software MATLAB via the thin operation of the bwmorph function. Each subiteration starts by examining the neighborhood of every pixel in the binary image, and on the basis of a point set of pixel-deletion criteria, it decides whether the pixel can be withdraw or not. These subiterations goes on until no more pixels can be removed.Figure 4. 2 (a)Original Image, (b)Enhanced Image, (c)Binarised Image, (d)Thinned Image Chapter 5 Feature Extraction After improving quality of the ? ngerprint image we extract features from binarised and change state images. We extract address point, minutiae and key(used for one to many matching). 5. 1 revealing the Reference Point Reference point is very important feature in advanced matching algorithms because it provides the location of origin for marking minutiae. We ? nd the interview point using the algorithm as in 2. Then we ? nd the relative position of minutiae and estimate the orientation ? ld of the reference point or the singular point. The technique is to extract meaning and delta points using Poincare Index. The value of Poincare index is 180o , ? 180o and 0o for a core, a delta and an ordinary point respectively. Complex ? lters are used to produce blur at di? erent clotures. Singular point (SP) or reference point is the point of maximum ? lter response of these ? lters applied on image. Complex ? lters , exp(im? ) , of order m (= 1 and -1) are used to produce ? lter response. Four level resolutions are used herelevel 0, level 1, level 2, level 3.Level 3 is lowest resolution and level 0 is highest resolution. Only ? lters of ? rst order are used h = (x + iy)m g(x, y) where g(x,y) is a Gaussian de? ned as g(x, y) = exp? ((x2 + y 2 )/2? 2 ) and m = 1, ? 1. Filters are applied to the complex valued orientation tensor ? eld image z(x, y) = (fx + ify )2 and not directly to the image. here(predicate) f x is the derivative of the original image in the x-direction and f y is the derivative in the y-direction. To ? nd the position of a possible 35 36 CHAPTER 5. FEATURE fall Figure 5. 1 Row 1 ? lter response c1k , k = 3, 2, and 1. Row 2 ? ter response c2k , k = 3, 2, and 1. SP in a ? ngerprint the maximum ? lter response is extracted in image c13 and in c23 (i. e. ?lter response at m = 1 and level 3 (c13 ) and at m = ? 1 and level 3 (c23 )). The anticipate is done in a window computed in the preliminary higher(prenominal) level (low resolution). The ? lter response at lower level (high resolution) is used for ? nding response at higher level (low resolution). At a certain resolution (level k), if cnk (xj , yj ) is higher than a threshold an SP is found and its position (xj , yj ) and the complex ? lter response cnk (xj , yj ) are noted. 5. 2 5. 2. 1Minutiae Extraction and Post-Processing Minutiae Extraction The most commonly utilise method of minutiae extraction is the Crossing Number (CN) concept 1 . This method involves the use of the skeleton image where the ridge ? ow pattern is eight-connected. The minutiae are extracted by scan the local neighborhood of each ridge pixel in the image using a 3 x 3 window. The CN value is then computed, which is de? ned as half the sum of the di? erences between pairs of adjacent pixels in the eight-neighborhood. Using the properties of the CN as shown in ? gure 5, the ridge pixel can then be classi? d as a ridge ending, bifurcation or non-minutiae point. For example, a ridge pixel with a CN of one corresponds to a ridge ending, and a CN of one-third corresponds to a bifurcation. 5. 2. MINUTIAE EXTRACTION AND POST-PROCESSING Table 5. 1 Properties of Crossing Number C N seat 0 Isolated Point 1 Ridge Ending Point 2 Continuing Ridge Point 3 Bifurcation Point 4 Crossing Point 37 Figure 5. 2 Examples of (a)ridge-ending (CN=1) and (b)bifurcation pixel (CN=3) 5. 2. 2 Post-Processing rancid minutiae may be introduced into the image due to factors such as noisy images, and image artefacts created by the thinning process.Hence, after the minutiae are extracted, it is necessary to employ a post-processing 1 stage in order to validate the minutiae. Figure 5. 3 illustrates some examples of false minutiae structures, which include the spur, hole, triangle and spike structures . It can be seen that the spur structure generates false ridge endings, where as both the hole and triangle structures generate false bifurcations. The spike structure creates a false bifurcation and a false ridge ending point. Figure 5. 3 Examples of typical false minutiae structures (c)Triangle, (d)Spike (a)Spur, (b)Hole, 38 CHAPTER 5.FEATURE EXTRACTION 5. 2. 3 Removing Boundary Min utiae For removing boundary minutiae, we used pixel- immersion approach. Any point on the boundary will have less white pixel parsimony in a window centered at it, as compared to inner minutiae. We calculated the limit, which indicated that pixel tautness less than that means it is a boundary minutiae. We calculated it according to following formula limit = ( w w ? (ridgedensity)) ? Wf req 2 (5. 1) where w is the window size, Wf req is the window size used to compute ridge density. Figure 5. 4 Skeleton of window centered at boundary minutiaeFigure 5. 5 intercellular substance Representation of boundary minutiae Now, in thinned image, we sum all the pixels in the window of size w centered at the boundary minutiae. If sum is less than limit, the minutiae is considered as boundary minutiae and is discarded. 5. 3. EXTRACTION OF THE KEY 39 5. 3 5. 3. 1 Extraction of the key What is key? Key is used as a hashing tool in this project. Key is small set of a few(prenominal) minutiae close st to reference point. We match minutiae sets, if the keys of savour and interview ? ngerprints matches. Keys are stored along with minutiae sets in the database.Advantage of using key is that, we do not perform honest matching every time for non-matching minutiae sets, as it would be time consuming. For large databases, if we go on matching mount minutiae set for every enrolled ? ngerprint, it would waste time unnecessarily. Two types of keys are proposed simple and complex. Simple key has been used in this project. Figure 5. 6 Key Representation Simple Key This type of key has been used in this project. Minutiae which constitute this key are ten minutiae closest to the reference point or centroid of all minutiae, in sorted 40 CHAPTER 5. FEATURE EXTRACTION order. Five ? lds are stored for each key value i. e. (x, y, ? , t, r). (x, y) is the location of minutiae, ? is the value of orientation of ridge related to minutia with respect to orientation of reference point, t is type of minutiae, and r is infinite of minutiae from origin. Due to inaccuracy and imperfection of reference point detection algorithm, we used centroid of all minutiae for construction of key. Complex Key The complex key stores more information and is structurally more complex. It stores vector of minutiae in which next minutiae is closest to previous minutiae, starting with reference point or centroid of all minutiae.It stores < x, y, ? , t, r, d, ? >. Here x,y,t,r,? are same, d is distance from previous minutiae entry and ? is di? erence in ridge orientation from previous minutiae. Data minutiaelist = Minutiae Set, refx = x-cordinate of centroid, refy = y-cordinate of centroid Result Key d(10)=null for j = 1 to 10 do for i = 1 to rows(minutiaelist) do d(i) Chapter 6 Partitioning of Database Before we partition the database, we perform gender estimation and classi? cation. 6. 1 Gender Estimation In 3, study on 100 staminates and 100 fe masculines revealed that signi? slope sex d i? erences occur in the ? ngerprint ridge density.Henceforth, gender of the candidate can be estimated on the basis of given ? ngerprint data. Henceforth, gender of the candidate can be estimated on the basis of given ? ngerprint data. establish on this estimation, distinct for a record in the database can be made faster. Method for ? nding mean ridge density and estimated gender The highest and lowest values for male and womanly ridge densities will be se twisted. If ridge density of query ? ngerprint is less than the lowest ridge density value of females, the query ? ngerprint is seemingly of a male. Similarly, if it is higher than highest ridge density value of males, the query ? gerprint is of a female. So the searching will be carried out in male or female domains. If the value is between these values, we search on the basis of whether the mean of these values is less than the density of query image or higher. 41 42 CHAPTER 6. PARTITIONING OF DATABASE Figure 6. 1 Gender Esti mation 6. 1. GENDER ESTIMATION Data Size of Database = N Ridge compactness of query ? ngerprint = s Result Estimated Gender i. e. male or female maleupperlimit=0 femalelowerlimit=20 mean=0 for image < femalelowerlimit then femalelowerlimit 43 if s < maleupperlimit then estimatedgender 44 CHAPTER 6.PARTITIONING OF DATABASE 6. 2 Classi? cation of Fingerprint We divide ? ngerprint into ? ve classes arch or tented arch, left loop, right loop, peal and unclassi? ed. The algorithm for classi? cation 4 is used in this project. They used a ridge classi? cation algorithm that involves three categories of ridge structuresnonrecurring ridges, type I recurring ridges and type II recurring ridges. N1 and N2 represent number of type I recurring ridges and type II recurring ridges respectively. Nc and Nd are number of core and delta in the ? ngerprint. To ? nd core and delta, separate 135o blocks from orientation image. 35o blocks are shown in following ? gures. Figure 6. 2 135o blocks of a ? ngerprint Based on number of such blocks and their relative positions, the core and delta are found using Poincare index method. After these, classi? cation is done as following 1. If (N2 > 0) and (Nc = 2) and (Nd = 2), then a whorl is identi? ed. 2. If (N1 = 0) and (N2 = 0) and (Nc = 0) and (Nd = 0), then an arch is identi? ed. 3. If (N1 > 0) and (N2 = 0) and (Nc = 1) and (Nd = 1), then categorize the input using the core and delta mind algorithm4. 4. If (N2 > T2) and (Nc > 0), then a whorl is identi? ed. 5.If (N1 > T1) and (N2 = 0) and (Nc = 1) then classify the input using the core and delta assessment algorithm4. 6. If (Nc = 2), then a whorl is identi? ed. 7. If (Nc = 1) and (Nd = 1), then classify the input using the core and delta assessment algorithm4. 8. If (N1 > 0) and (Nc = 1), then classify the input using the core and delta assessment algorithm. 6. 3. PARTITIONING 9. If (Nc = 0) and (Nd = 0), then an arch is identi? ed. 10. If none of the above condition s is satis? ed, then egest the ? ngerprint. 45 Figure 6. 3 Fingerprint Classes (a)Left Loop, (b)Right Loop, (c)Whorl, (d1)Arch, (d2)Tented Arch . 3 Partitioning After we estimate gender and ? nd the class of ? ngerprint, we know which ? ngerprints to be searched in the database. We nearly divide database into one-tenth using the above parameters. This would roughly reduce identi? cation time to one-tenth. 46 CHAPTER 6. PARTITIONING OF DATABASE Figure 6. 4 Partitioning Database Chapter 7 Matching Matching means ? nding most appropriate like ? ngerprint to query ? ngerprint. Fingerprints are matched by matching set of minutiae extracted. Minutiae sets never match completely, so we compute match score of matching. If match score satis? s accuracy needs, we call it lucky matching. We used a new key based one to many matching intended for large databases. 7. 1 Alignment Before we go for matching, minutiae set need to be organizeed(registered) with each other. For alignment problems, we used hough transform based registration technique connatural to one used by Ratha et al5. Minutiae alignment is done in two steps minutiae registration and pairing. Minutiae registration involves aligning minutiae using parameters < ? x, ? y, ? > which range within speci? ed limits. (? x, ? y) are translational parameters and ? is rotational parameter.Using these parameters, minutiae sets are rotated and translated within parameters limits. Then we ? nd pairing scores of each chemise and displacement giving maximum score is registered as alignment transformation. Using this transformation < ? x, ? y, ? >, we align query minutiae set with the database minutiae set. Algorithm is same as in 5 but we have excluded factor ? s i. e. the scaling parameter because it does not a? ect much the alignment process. ? lies from -20 degrees to 20 degrees in steps of 1 or 2 generalized as < ? 1 , ? 2 , ? 3 ? k > where k is number of rotations applied.For every query minutiae i we check if ? k + ? i = ? j where ? i and ? j are orientation 47 48 CHAPTER 7. co-ordinated parameters of ith minutia of query minutiae set and j th minutia of database minutiae set. If condition is satis? ed, A(i,j,k) is ? agged as 1 else 0. For all these ? agged values, (? x, ? y) is calculated using following formula ? (? x , ? y ) = qj ? ? cos? sin? ? ? ? pi , (7. 1) ?sin? cos? where qj and pi are the coordinates of j th minutiae of database minutiae set and ith minutiae of query minutiae set respectively. Using these < ? x, ?y, ? k > values, whole query minutiae set is aligned.This aligned minutiae set is used to compute pairing score. Two minutiae are said to be paired only when they lie in same bounding box and have same orientation. Pairing score is (number of paired minutiae)/(total number of minutiae). The i,j,k values which have highest pairing score are ? nally used to align minutiae set. Co-ordinates of aligned minutiae are found using the formula ? qj = ? cos? sin? ? ? ? pi + (? x , ? y ), (7. 2) ?sin? cos? After alignment, minutiae are stored in sorted order of their distance from their centroid or core. 7. 2 Existing Matching TechniquesMost popular matching technique of today is the simple disposed(p) n2 matching where n is number of minutiae. In this matching each minutiae of query ? ngerprint is matched with n minutiae of sample ? ngerprint giving total number of n2 comparisons. This matching is very orthodox and gives headache when identi? cation is done on large databases. 7. 3 One to Many matching Few algorithms are proposed by many researchers around the world which are better than normal n2 matching. But all of them are one to one veri? cation or one to one identi? cation matching types. We developed a one to many matching technique which uses key as the hashing tool.Initially, we do not match minutiae sets instead we per- 7. 3. ONE TO MANY MATCHING 49 form key matching with many keys of database. Those database ? ngerprints whose keys match with key of query ? ngerprint, are allowed for broad(a) minutiae matching. Key matching and rise matching are performed using k*n matching algorithm discussed in later section. Following section gives method for one to many matching. Data Query Fingerprint Result Matching Results Acquire Fingerprint, Perform Enhancement, Find Fingerprint Class, Extract Minutiae, Remove Spurious and Boundary Minutiae, Extract Key,Estimate Gender M . 3. 1 Method of One to Many Matching The matching algorithm will be involving matching the key of the query ? ngerprint with the many(M) keys of the database. Those which matches ,their full matching will be processed, else the query key will be matched with next M keys and so on. 50 Data Gender, Class, i Result Matching Results egender CHAPTER 7. MATCHING if keymatchstatus = success then eminutiae 7. 4 Performing key match and full matching Both key matching and full matching are performed using our k*n matching technique. Here k is a immutable( recommended value is 15) chosen by us.In this method, we match ith minutiae of query set with k leftover minutiae of sample set. Both the query sets and sample sets must be in sorted order of distance from reference point or centroid. ith minutia of query minutiae list is matched with top k unmatched minutiae of database minutiae set. This type of matching reduces matching time of n2 to k*n. If minutiae are 80 in number and we chose k to be 15, the total number of comparisons will reduce from 80*80=6400 to 80*15=1200. And this means our matching will be k/n times faster than n2 matching. 7. 5. conviction COMPLEXITY OF THIS MATCHING TECHNIQUE 51 Figure 7. One to Many Matching 7. 5 Time Complexity of this matching technique let s = size of the key, n = number of minutiae, N = number of ? ngerprints matched till successful identi? cation, k = constant (see previous section). There would be N-1 frustrated key matches, one successful key match, one successful full match. Time for N-1 unsuccessful key matches is (N-1)*s*k (in worst case), for successful full match is s*k and for full match is n*k. Total time is (N-1)*s*k+n*k+s*k = k(s*N+n). Here s=10 and we have minify database to be searched to 1/10th ,so N matching technique, it would have been O(Nn2 ).For large databases, our matching technique is best to use. Averaging for every ? ngerprint, we have O(1+n/N) in this identi? cation process which comes to O(1) when N >> n. So we can say that our identi? cation system has constant average matching time when database size is millions. Chapter 8 Experimental Analysis 8. 1 Implementation Environment We tested our algorithm on several(prenominal) databases like FVC2004, FVC2000 and Veri? nger databases. We used a computer with 2GB RAM and 1. 83 gigahertz Intel Core2Duo processor and softwares like Matlab10 and MS access10. 8. 2 8. 2. 1 Fingerprint Enhancement Segmentation and NormalizationSegmentation was performed and it generated a mask matrix which has valu es as 1 for ridges and 0 for background . Normalization was done with mean = 0 and variance = 1 (? g 8. 1). Figure 8. 1 Normalized Image 52 8. 2. FINGERPRINT ENHANCEMENT 53 8. 2. 2 Orientation Estimation In orientation estimation, we used block size = 3*3. Orientations are shown in ? gure 8. 2. Figure 8. 2 Orientation Image 8. 2. 3 Ridge Frequency Estimation Ridge density and mean ridge density were calculated. Darker blocks indicated low ridge density and vice-versa. Ridge frequencies are shown in ? gure 8. 3. Figure 8. 3 Ridge Frequency Image 8. 2. 4Gabor Filters Gabor ? lters were diligent to enhance quality of image. Orientation estimation and ridge frequency images are requirements for implementing gabor ? lters. ?x and ? y are taken 0. 5 in Raymond Thai, but we used ? x = 0. 7 and ? y = 0. 7. Based on these values , we got results which were satis? able and are shown in ? gure 8. 4. 54 CHAPTER 8. observational outline Figure 8. 4 Left-Original Image, Right-Enhanced Image 8. 2. 5 Binarisation and Thinning After the ? ngerprint image is enhanced, it is then converted to binary form, and submitted to the thinning algorithm which reduces the ridge thickness to one pixel wide.Results of binarisation are shown in ? gure 8. 5 and of thinning are shown in ? gure 8. 6. Figure 8. 5 Binarised Image 8. 3. FEATURE EXTRACTION 55 Figure 8. 6 Thinned Image 8. 3 8. 3. 1 Feature Extraction Minutiae Extraction and Post Processing Minutiae Extraction Using the crossing number method, we extracted minutiae. For this we used skeleton image or the thinned image. Due to low quality of ? ngerprint, a lot of false and boundary minutiae were found. So we moved forward for post-processing step. Results are shown in ? gure 8. 7 and 8. 8. Figure 8. 7 All Extracted Minutiae 56 CHAPTER 8. EXPERIMENTAL ANALYSISFigure 8. 8 Composite Image with spurious and boundary minutiae After Removing Spurious and Boundary Minutiae False minutiae were removed using method described in earlier sect ion. For removing boundary minutiae, we employed our algorithm which worked ? ne and minutiae extraction results are shown in table 8. 2. Results are shown in ? gure 8. 9 and 8. 10. Figure 8. 9 Minutiae Image after post-processing As we can see from table 8. 2 that removing boundary minutiae considerably reduce the number of false minutiae from minutiae extraction results. 8. 4. GENDER ESTIMATION AND CLASSIFICATION 57 Figure 8. 0 Composite Image after post-processing Table 8. 1 Average Number of Minutiae before and after post-processing DB After After Removing After Removing Used Extraction Spurious Ones Boundary Minutiae FVC2004DB4 218 186 93 FVC2004DB3 222 196 55 8. 3. 2 Reference Point Detection For reference point extraction we used complex ? lters as described earlier. For a database size of 300, reference point was found with success rate of 67. 66 percent. 8. 4 8. 4. 1 Gender Estimation and Classi? cation Gender Estimation Average ridge density was calculated along with mini mum and maximum ridge densities shown in table 8. . Mean ridge density was used to divide the database into two parts. This trim back database size to be searched by half. Based on the information available about the gender of enrolled student, we can apply our gender estimation algorithm which will further increase the speed of identi? cation. 8. 4. 2 Classi? cation Fingerprint classi? cation was performed on both original and enhanced images. Results were more accurate on the enhanced image. We used same algorithm as in sec 6. 2 to classify the ? ngerprint into ? ve classes arch, left loop, right loop, whorl and 58 CHAPTER 8.EXPERIMENTAL ANALYSIS Figure 8. 11 Plotted Minutiae with Reference Point(Black Spot) Table 8. 2 Ridge absorption Calculation Results Window Minimum Maximum Mean Total Average Size Ridge Ridge Ridge Time Time taken Density Density Density Taken Taken 36 6. 25 9. 50 7. 87 193. 76 sec 1. 46 sec unclassi? ed. This classi? cation was used to divide the database into ? ve parts which would reduce the database to be searched to one-? fth and ultimately making this identi? cation process ? ve times faster. Results of classi? cation are shown in table 8. 4, 8. 5 and 8. 6. 8. 5 EnrollingAt the time of enrolling personal details like name, semester, gender, age, roll number etc. were asked to input by the user and following features of ? ngerprint were saved in the database (1)Minutiae Set (2)Key (3)Ridge Density (4)Class Total and average time taken for enrolling ? ngerprints in database is shown in table 8. 6. MATCHING Table 8. 3 Classi? cation Results on Original Image Class No. of (1-5) Images 1 2 2 2 3 3 4 4 5 121 Table 8. 4 Classi? cation Results on Enhanced Image Class No. of (1-5) Images 1 8 2 3 3 3 4 6 5 112 59 8. 7. All the personal details were stored in the MS annoy database and were modi? d by running sql queries inside matlab. Fingerprint features were stored in txt format inside a separate folder. When txt ? le were used, the pr ocess of enrolling was faster as compared to storing the values in MS Access DB. It was due to the overhead of connections, running sql queries for MS Access DB. 8. 6 Matching Fingerprint matching is required by both veri? cation and identi? cation processes. 8. 6. 1 Fingerprint Veri? cation Results Fingerprint veri? cation is the process of matching two ? ngerprints against each other to verify whether they belong to same person or not. When a ? gerprint matches with the ? ngerprint of same individual, we call it true accept or if it doesnt, we call it false reject. In the same way if the ? ngerprint of di? erent individuals match, we call it a false accept or if it rejects them, it is true reject. False Accept Rate ( faraway) and False lour Rate (FRR) are the error rates which are used to express matching trustability. FAR is de? ned by the formula 60 CHAPTER 8. EXPERIMENTAL ANALYSIS Table 8. 5 Time taken for Classi? cation Image Average Total Taken Time(sec) Time(sec) Original 0. 5233 69. 07 Enhanced 0. 8891 117. 36 Table 8. Time taken for Enrolling No. of Storage Average Total Images Type Time(sec) Time(hrs) 294 MS Access DB 24. 55 2. 046 60 MS Access DB 29. 37 0. 49 150 TXT ? les 15. 06 1. 255 F AR = FA ? 100, N (8. 1) FA = Number of False Accepts, N = Total number of veri? cations FRR is de? ned by the formula FR ? 100, N F RR = (8. 2) FR = Number of False Rejects. FAR and FRR calculated over six templates of Veri? nger DB are shown in table 8. 8. This process took nearly 7 hours. 8. 6. 2 Identi? cation Results and Comparison with Other Matching techniques Fingerprint identi? cation is the process of identifying a query ? gerprint from a set of enrolled ? ngerprints. Identi? cation is usually a pokey process because we have to search over a large database. currently we match minutiae set of query ? ngerprint with the minutiae sets of enrolled ? ngerprints. In this project, we store key in the database at the time of enrolling. This key as explaine d in sec 5. 3 helps in 8. 6. MATCHING Table 8. 7 Error Rates FAR FRR 4. 56 12. 5 14. 72 4. 02 61 Figure 8. 12 Graph Time taken for Identi? cation vs Size of Database(key based one to many identi? cation) reducing matching time over non-matching ? ngerprints. For non-matching enrolled ? gerprints, we dont perform full matching, instead a key matching. Among one or many keys which matched in one iteration of one to many matching, we allow full minutiae set matching. Then if any full matching succeeds, we perform post matching steps. This identi? cation scheme has lesser time complexity as compared to conventional n2 one to one identi? cation. Identi? cation results are shown in table 8. 9. The graph of time versus N is shown in ? gure 8. 13. Here N is the index of ? ngerprint to be identi? ed from a set of enrolled ? ngerprints. Size of database of enrolled ? ngerprints was 150. So N can vary from