You are on page 1of 5

International Journal of Engineering Research & Technology (IJERT

)
ISSN: 2278-0181
Vol. 2 Issue 10, October - 2013

Finger Print Matching Algorithm for Android
Authors :

Kavita Rathi
P G Student Datta Meghe College of Engg.
Airoli ,Navi Mumbai.400708

Sudhir Sawarkar
Principle of Datta Meghe College of Engg.

Abstract able to store large amounts of information. In addition,
applications of these mobile devices have been
RT
We propose a secure robust, and low cost biometric broadened to include services requiring security
authentication system on the mobile devices. Mobile functions, e-banking, e-commerce, etc. These have
computing devices such as Smartphone’s, tablets and caused mobile-device manufacturers to install more
IJE

wireless enabled laptops are getting widely popular and secure authentication methods than the simple
being frequently used to access information in secure conventional personal-identification-number-based
networks through third party providers. With the method. To satisfy these demands, some contemporary
increasing number of intrusion, device and identity theft mobile devices adapt the fingerprint-verification
incidents, improved security measures are becoming function. These devices embed extra sensors to capture
essential for accessing devices as well as various mobile fingerprint images at low costs. However, additional
applications. Hence security is of paramount importance sensor increases the cost of the mobile device because
in today’s world. The project we propose to make will new product design and platform development are
firstly need the user to register with application the user additionally required. To reduce the costs, built-in
will have to provide credentials along with his cameras can be used to obtain the fingerprint images.
fingerprints and the fingerprint of the member to grant This approach is very attractive because additional
his access to information. After registering with one just hardware costs are unnecessary, and modern cameras
has to authenticate using fingerprint if the fingerprint that are built in mobile devices can already provide high-
matched with that of the stored in database the required resolution and close-up images. Once a built-in camera
information is displayed in this way, the user need not is used as the input device, the acquisition of suitable
remember any password in order to log in to images for fingerprint verification is the fundamental
application. This will reduce the cumbersome task of step. The prevailing techniques of user authentication,
remembering the password and increase the security as which involve the use of either and user IDs (identifiers),
the fingerprint is unique. The proposed architecture is or identification cards and PINs (personal identification
expected to provide secure access to only legitimate numbers), suffer from several limitations. User IDs and
users of a system within a short period of time and passwords can be changed as often as required. Yet, the
prohibit intruders from gaining any access. user only has a limited number of biometric features
Keywords - biometrics, information (one face, ten fingers, and two eyes).And which cannot
security, mobile computing, systems flowchart be copied. Additionally, a biometric property of an
individual can be lost only in case of serious accident.
1. Introduction The aim of this dissertation is to develop a secure
Recently, performance of mobile devices, such as architecture for mobile network using inbuilt camera.
cellular phones and personal digital assistants (PDAs), First, study several biometric identification techniques
has improved dramatically the. Mobile devices are now are reviewed and the security challenges with each
technique is investigated. Then a generic securely

IJERTV2IS101166 www.ijert.org 3819

org 3820 . used by 71% of the mobile developer Figure1:Atrix4G with fingerprint scanner. and using android. digital cameras and other electronics. software. Android's open nature has further encouraged a large community of developers and enthusiasts to use the open-source code as a foundation for community-driven projects. Inc. population. These factors have contributed towards The objective of this dissertation is to develop a secure making Android the world's most widely used architecture for mobile network and system involving Smartphone platform.000 apps available for Android. released running other operating systems.ijert. As a IJE biometrics is proposed and evaluated. Apache License. challenges with each technique is investigated. it has seen additional applications on televisions.5). If the library finds Intel's Integrated Performance Primitives on the system. The first Android- advance version this will work with powered phone was sold in October 2008. October .2013 approach for mobile devices using biometrics is along with the founding of the Open Handset Alliance: a proposed and evaluated. A developer survey conducted in April–May 2013 found that Android is the most popular platform for developers. several biometric quarter of 2010. Android has a large community of developers writing applications ("apps") that extend the functionality of devices. it will use these proprietary optimized routines to accelerate itself. which Google backed financially and later bought in 2005. and open CV library so that it will work telecommunication companies devoted to advancing on all android operating systems mobile version 2. It focuses mainly on real-time image processing.. In October 2012. but it still retains a less comprehensive though extensive Figure 1: android older C interface. written primarily in a customized version of the Java programming language. Open CV is written in C++ and its primary interface is in C++. Additionally. and the estimated number of applications downloaded from Google Play. Java and MATLAB/OCTAVE (as of version primarily for touch-screen mobile devices such as 2. Android was unveiled in 2007 IJERTV2IS101166 www. lightweight operating system for high tech generic securely approach for mobile devices using devices without developing one from scratch. by Android. and now supported by Willow Garage and Itseez. developed by Intel. Then a customizable. The API for these interfaces can be found in the Smartphone’s and tablet computers. First. and the software of choice for RT identification techniques are reviewed and the security technology companies who require a low-cost. Android's primary app store. This open-source code and permissive licensing allows the software to be freely modified and distributed by device manufacturers. This project is implemented by consortium of hardware. The library is cross-platform. International Journal of Engineering Research & Technology (IJERT) ISSN: 2278-0181 Vol. despite being primarily designed for phones and tablets. was 25 billion. result. Open CV: Open CV (Open Source Computer Vision Library) is a library of programming functions mainly aimed at real- time computer vision. 2 Issue 10. which add new features for advanced users or bring Android to devices which were Android OS: officially. wireless carriers and enthusiast developers.2 to open standards for mobile devices. there were approximately 700. games consoles. overtaking Symbian in the fourth biometric information. Initially developed online documentation. There are now full interfaces in Android is a Linux-based operating system designed Python. Android is all camera starting from 2megapixel it does not need a open source and Google releases the code under the very high resolution camera.

Hence a typing pattern of different words is b) ear. When smart phone and tablet users type various words in their phone using keypad or touch screen.3 strength and weakness and the choice is depend on the Skin color segmentation: In isolating skin color region user. breaks and vibration frequencies. time between key strokes are recorded and Figure 4. each pixel is classified as skin or Different biometric signatures are shown in fig4. The camera.org 3821 . can be used to more precisely identify individuals [4]. Based on the color depending upon on the operation of the application. 2 Issue 10. c) face. HIS and YCrCb color space specific biometric and an application is determined transformations are commonly used. The match between a normalized RGB. Examples of biometric characteristics: a) DNA. Behavioral traits. Majority of the phones available in market do not have any dedicated component for recording biometric data. It can also be noted that perpetrators can use recorded voice and try to gain access as the legitimate user. Face recognition techniques have also been applied to identify individuals using mobile phone cameras [3]. Review Of literature in recent years. Eye features.from mobile phone cameras and extract fingerprint features for scanning and verification purposes [2]. analyzed. especially after emergence of smart phones with higher processing capabilities. j) iris. It is also possible to enhance fingerprint images and the old Fujitsu F505i [7]. teeth shape authentication can make the recognition performance more accurate where individual and overall teeth shapes RT are analyzed to perform matching [5]. microphone and touch screen are mainly used as interfaces to collect biometric data. g) fingerprint. such as keystroke dynamics. h) gait. f)hand vein. where from voice traits. the user might be asked to speak a selected sentence or words. International Journal of Engineering Research & Technology (IJERT) ISSN: 2278-0181 Vol. In IJE this case. eye shape and distance between eyes. In 2011. Motorola released Atrix 4G smart phone which includes a dedicated fingerprint reader. are another set of features that can be used to authenticate users [7]. m) signature. where the lines of the palm are analyzed for identification. no biometric is “best”. In addition. k) palm print. III. e) hand thermo derived which can then be utilized to verify the identity gram. the smart phone user would be asked to speak a set of words or sentences and thus the response are recorded to retrieve voice samples.[1] component values. cold or emotional states. pupil. tones. face recognition can be performed from pictures taken with moderate lighting conditions. Biometric has its mobile security is shown in the block diagram of Fig. geometry. can be verified. finger lines also have the potential for persona l identification for mobile phones.2013 II. d) facial thermo gram.The use non-skin. and n) voice. In this architecture RGB matching algorithm is of biometrics to secure mobile devices has been explored used the number of pixels used to represent the finger in mobile phone photos dominates the pixels used to IJERTV2IS101166 www.Like palm print. As most smart phones and tablets are now equipped with medium to high resolution cameras. For authentication purpose. October . such as pitch magnitudes. Authentication using voice may fail to identify individuals due to factors such as flue. such as iris scan. l) retina. Proposed architecture Different numbers of biometric signatures are present An overview of the proposed system architecture for and are used in different devices.ijert. i) hand of a person [8].Voice or speech recognition using microphone can also be used for user authentication in order to access mobile devices [6].

ijert. 2 Issue 10. continue. Therefore these  If the Euclidian distance between the approaches are likely to give good results on handsets. namely R G B. 3) Read the reference image from the database. continue. simplest methods to implement on a mobile phone could 10) Stop.  If the Euclidian distance between the planes does not lie within the threshold. planes is within To reduce the effects of illumination variations scale-by.  If the Euclidian distance between the max color balancing is done. go to step 10 8) Compare the B plane of the reference image with the B plane of the current image IJERTV2IS101166 www. continue. It should RT necessarily be in Bitmap format. go to step 10 7) Compare the R plane of the reference image with the R plane of the current image  If the Euclidian distance between the planes is within the threshold. namely R G B. be do the calculations in RGB space Figure5: An overview of the proposed mobile device security architecture Algorithm: 1) Read the image from the camera. International Journal of Engineering Research & Technology (IJERT) ISSN: 2278-0181 Vol.2013 represent other part of the body. October . go to step 10 images in extreme illumination conditions. IJE 4) Split the reference image into three bit planes. Many complex methods planes does not lie within the have been used in the literature to do the processing of threshold. go to step 10 6) Compare the G plane of the reference image with the G plane of the current image  If the Euclidian distance between the planes is within the threshold.org 3822 .  If the Euclidian distance between the planes does not lie within the threshold. 2) Split the current image into three bit planes. 5) Compare the R plane of the reference image with the R plane of current image  If the Euclidian distance between the planes is within the threshold.  If the Euclidian distance between the planes does not lie within the threshold. One of the 9) The images have successfully matched.

March. devices using biometric. Choi. "An introduction to Now days tshe Smartphone’s stored our important biometric. In this project we are taking no of selection for fingerprint recognition with a mobile- samples and by using RGB algorithm the samples are device matched with the samples stored in database. S. November. edge detection. Jeong. Choi. Byun. like KMEANS. "A study on iris localization and recognition on mobile phones. NY. Toh and H. Lee. November 2010. Choi and J. P. not as easy as it might appear. no. Maiorana. Man. This implemented architecture "Keystroke dynamics authentication for mobile phones. K. Kang. International Journal of Engineering Research & Technology (IJERT) ISSN: 2278-0181 Vol." IJE Wrong fingerprint: IEEE Transaction on Consumer Electronics. pp. [7] A. B. Lee and D. González-Carballo and to steal. Our aim if the costly mobile device is IEEE Transaction On Circuits and System for Video stolen by theft it should not be usable for the theft." Pattern Recogn. vol. future we will use different image processing algorithm issue 1. vol. The word biometrics is very often used as a synonym for the perfect security without need of dedicated scanner and the security is increased against intrusion theft and manipulation of devices. M. no. samples Detected [4] K. "Recognizable- using biometric. It is much secured and difficult to steal image as biometric is unique. Park. "Modeling Person1 100 33 77% cryptographic properties of voice and voice-based entity authentication. In vol. Campisi. Jain. vol. Cochinwala and H. and M. It is very secured and difficult [8] E. January 2008. No of Correctly Accuracy 44. Neri. Wilkerson. and Cybernetics Society. pp." Proceedings of the 2007 ACM workshop Person2 100 32 78% on Digital identity management (DIM '07). K. Chung and K. 2011 secure biometric systems is. 386 – 400. E. 14. no. In this project a robust system is developed for A. 77%-78% accuracy is possible with RGB algorithm. Initially camera. [3] K. "Realtime training on mobile Correct fingerprint: devices for face recognition applications. 2008. 40 – 43. Signal Process. Crescenzo.” IEEE Security & Privacy. 4 – 20. pp.2013 References: Result and Discussion: [1] A. pp. however. [6] G. IJERTV2IS101166 www. vol. mobile devices security.ijert. Person2 100 72 72% [5] D. Adv. pp." Person1 100 82 82% EURASIP J. 1. 233 – 243.org 3823 ." information. Prabhakar. is very innovative idea to authenticate mobile devices [2] D. Peacock. 2011. pp. H. “Typing patterns: a key Conclusion: to user identification. K. X. February. Person3 100 34 76% USA. No of Wrongly Accuracy 6. This is very innovative idea to authenticate mobile 2004. Park. Even if the accuracy Proceedings of the 2011 ACM Symposium on Applied of the biometric techniques is not perfect yet making a Computing (SAC '11). H. Ross. Shim. February 2008. N. Kim. 53 – 61. Hong. 38. 21 – 26. 2004. This Technology. October . A. samples Detected 2678 – 2685. "Person RT authentication using Person3 100 79 79% face. 2007. teeth and voice modalities for mobile device security. New York. 56. 2 Issue 10." IEEE Systems. Kim. 2. Ke." can be used in all android mobile. pp.