Sunteți pe pagina 1din 3

Web Site: www.ijettcs.org Email: editor@ijettcs.org, editorijettcs@gmail.

com Volume 1, Issue 3, September October 2012 ISSN 2278-6856

Image Enhancement Techniques for Fingerprint Images


Annu Saini
Apaji Institute of Mathematics & Applied Computer Technology Department of computer Science and Electronics, Banasthali University, India

Abstract: Fingerprints are the most popular biometric


identification .Everyone is known to have unique fingerprints and these fingerprints patterns do not change throughout the life. This is the reason for the popularity of fingerprints as the biometric identifier. The quality of image plays very important role in matching the two fingerprints .But most of the fingerprint recognition systems results poor matching due to the poor quality of image. In this paper I have explained some image enhancement methods that will improve the quality of the image before matching them. These techniques help the fingerprint recognition systems to give better results.

Keywords: fingerprint, ridge, pattern, image, histogram, enhancement, thinning, binarisation, filtering.

1.2 Fingerprint features: The lines that show a pattern on the fingerprint are called Ridges and the space between two ridges is called a valley .Each fingerprint contains minutiae or ridge characteristics and the fingerprints are identified by examining these minutiae points .It is the minutiae points which give the property of uniqueness to the fingerprints . There are two types of minutiae points: ridge ending and bifurcation. Ridge ending represents the termination of a ridge. And bifurcation is the feature which shows the splitting of a ridge in two paths.

1. INTRODUCTION
Biometrics is the branch of science which refers to the identification of humans by their characteristics. As the biometric identifiers are unique to individuals, so they are more reliable in verifying identity of a human. The characteristics used for human identification on the basis of their universality, uniqueness, permanence, measurability, acceptability. There are various biometric identifiers but fingerprints are the most widely used among them. Among all other kind of popular personal identification methods, fingerprint identification is most reliable technique. Fingerprints have two properties that are used for identification: i) Uniqueness: This property says that everyone has unique fingerprints. No two persons even if they are twins can have same fingerprints. No two fingerprints on the same hand have ever been found to be alike .The pattern of ridges and valleys are different for every individual and for every finger. ii) Permanence: This property says that the fingerprints of a person are permanent i.e. they do not change throughout the lifetime of that person. Even if there is a scar in the finger of a person then the pattern of the ridges and valleys will recover after some time. 1.1 What are fingerprints? Fingerprint is an impression of the friction ridges of all or any part of the finger .Friction ridge is an area comprised of the combination of ridge flow, ridge characteristics, and ridge structure. Volume 1, Issue 3, September October 2012 Figure:1 (a)ridge ending and bifurcation (b) core and Delta There are also two other features which are also used for matching of fingerprints: Core and Delta. Core is supposed to be the center from where the pattern is made. Delta is the point on a ridge from where three patterns deviate. There are three main patterns of fingerprints: 1) Loop Pattern: The ridges flow in one side, recurve, touch or pass through an imaginary line drawn from the delta to the core and exit the pattern on the same side from which it entered. 2) Whorl Pattern: This pattern consists of a series of concentric circles. 3) Arch Pattern: In this pattern ridges flow in one side and flow out the opposite side. There are no deltas in an arch pattern.

Figure 2: (a)Arch Pattern(b)Loop Pattern(c)Whorl Pattern

Page 215

Web Site: www.ijettcs.org Email: editor@ijettcs.org, editorijettcs@gmail.com Volume 1, Issue 3, September October 2012 ISSN 2278-6856
1.3 Fingerprint recognition Fingerprint recognition refers to the matching of two fingerprints using an automated system .The ridges pattern are used to match the fingerprints of two persons .and the minutiae are used for this matching .the minutiae points of two sample fingerprints are matched and then verified as they matched or not. The fingerprint recognition works on two domains: verification and identification. In fingerprint verification there is one to one match of fingerprints .And in identification there is one to many match for fingerprints. .So the fingerprints must be enhanced to get all the features of the fingerprints. There are many techniques that help in enhancing the image of fingerprint which are explained in this paper: 1) Histogram Modeling :The histogram of an image represents the relative frequency of the various gray levels of an image .Using histogram equalization we can get a uniform histogram for the output image .Thus this is a technique for the improvement in the contrast of an image by adjusting the intensity of every gray level of the image. In this technique without affecting the global contrast the lower contrast areas gain a higher contrast .The results of histogram equalization are shown in the following figure :
original image histogram equilized image

2. RELATED WORK
L.Hong et al. [1] present a fast fingerprint enhancement algorithm, which was based on local ridge orientation and frequency and which as a result improved the clarity of ridge and valley structure of the input fingerprint image. Sharat S. Chikkerur et al. [2] introduced an approach for fingerprint enhancement based on Short Time Fourier transform (STFT) Analysis algorithm proposed by them estimated all the intrinsic properties of fingerprints .also they compared their results with other filtering techniques and showed improvements in their results. M. Sepasian et al. [3] investigated the performance of a three step procedure for the fingerprint identification and enhancement using CLAHE (contrast limited adaptive histogram equalization) with clip limit, standard deviation and sliding neighborhood as stages during processing of fingerprint image. S. Greenberg et al. [4] proposed two methods for image enhancement. One is using histogram equalization, Wiener filtering and image binarization and other is using anisotropic filter for direct gray scale enhancement .and compared their results with other methods .Y. He et al. [5] developed a fingerprint image enhancement algorithm based on orientation fields. Tsai-Yang Jea et al. [6] presented an approach that use localized secondary features derived from relative minutiae information .A flow network based matching technique was introduced to obtain one to one correspondence of those features. D.K. Misra et al. [7] proposed a method for fingerprint enhancement in Fourier domain and matching based on minutiae matching. Jianwei Yang et al. [8] presented a fingerprint feature extraction method through which minutiae are extracted directly from original gray-level fingerprint images without binarization and thinning.

x 10 3 2.5 2 1.5 1 0.5 0 0

original hist ogram image 3.5 3 2.5 2 1.5 1 0.5 0 50 100 150 200 250

x 10

equalized histogram image

50

100

150

200

250

Figure 3: (a) original image (b) histogram equalized image(c) histogram of original image (d) histogram of equalized image 2) Various filtering methods: Median Filtering: Median filtering is a nonlinear filtering process used to remove impulsive or "salt & pepper" type noise. In median filtering [10] the input pixel is replaced by the median of the pixels contained in a window around the pixel, V (m, n) =median{y (m-k, n-l), (k, l) Where W is a suitably chosen window .The algorithm for median filtering requires arranging the pixel values of the window in an order and then choosing the median value among those pixels.

3. IMAGE ENHANCEMENT TECHNIQUES


The image acquired for the matching must be of good quality and it must be free of any type of noise. The less the noise in the fingerprint image the more accurate is the matching. So a good quality fingerprint image is desirable for the improved results of the matching. However it is not always easy to obtain a good quality of fingerprint as they may be corrupted due to the skin problems, scars on fingers .Also because of these cuts and scars a good quality fingerprint image may give false minutiae results Volume 1, Issue 3, September October 2012

Figure 4: results of applying median filter to an image High Pass filtering: When we want to extract the edges of the images we use high pass filters .High pass filters helps to sharpen the edges of the image. For this purpose a fraction of the high pass filtered image is added to the original image .The high pass filtered image is obtained by subtracting the blurred or low pass filtered image from the original image. Page 216

Web Site: www.ijettcs.org Email: editor@ijettcs.org, editorijettcs@gmail.com Volume 1, Issue 3, September October 2012 ISSN 2278-6856
Weiner Filtering: This method restores the image even in the presence of blur as well as noise. It uses a pixel wise adaptive method for noise reduction. affected by the noise, blur etc. So there is a need for enhancing the fingerprint images to get better results .The techniques presented in this paper will help in enhancing the fingerprint image and help in better matching in fingerprint recognition systems .

REFERENCES
[1] L.Hong, Y.Wan, and A.K Jain Fingerprint Image Enhancement: Algorithm and Performance Evaluation in IEEE Transactions on pattern analysis and Machine intelligence, Vol. 20, issue 8, august 1998 [2] Sharat S. Chikkerur, Alexander N. Cartwright, V. Govindaraju Fingerprint Image Enhancement using STFT Analysis submitted at Center for Unified Biometrics and Sensors. [3] M. Sepasian, W. Balachandran and C. Mares Image Enhancement for Fingerprint Minutiae-Based Algorithms Using CLAHE, Standard Deviation Analysis and Sliding Neighborhood in Proceedings of the World Congress on Engineering and Computer Science 2008 San Francisco, USA. [4] S. Greenberg, M. Aladjem, D. Kogan, I. Dimitrov Fingerprint Image Enhancement using Filtering Techniques submitted at Electrical and Computer Engineering Department, Ben-Gurion University of Negev, Israel. [5] Y. He , J. Tian , X. Luo, T. Zhang Image enhancement and minutiae matching in fingerprint verification in Pattern recognition letters 24(2003) 1349-1360. [6] Tsai-Yang Jea, V. Govindaraju A minutia-based partial fingerprint recognition system in Pattern Recognition 38 (2005) 1672 1684. [7] D.K. Misra, Dr.S.P. Tripathi and A. Singh Fingerprint Image Enhancement, Thinning and Matching in International Journal of Emerging Trends & Technology in Computer Science Volume 1, Issue 2, July August 2012. [8] Jianwei Yang, Lifeng Liu, and Tianzi Jiang* An Improved Method for Extraction of Fingerprint Featuresat National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences. [9]A.K. Jain, A. Fundamentals of Digital Image Processing; Prentice Hall, 1989. *figures are the results of coding applied to FVC2004 image database. AUTHOR Annu Saini received the M.Sc. degree in computer science from Banasthali University in 2011 and pursuing M.Tech in computer science from Banasthali university (2011-13).Now she is doing research in the field of image processing as a part for completing her degree. Page 217

Figure 5: results of applying Weiner filter to an image Gabor Filtering: Gabor filter is a linear filter whose impulse response is defined by a harmonic function multiplied by a Gaussian function .Gabor filters optimally capture both local orientation and frequency information from a fingerprint image. Once the ridge orientation and ridge frequency are determined, then they are used to construct the Gabor filter. In fingerprint enhancement, Gabor filter can be tuned to specific frequency and orientation values. Gabor filter can enhance the ridges in the direction of local orientation effectively preserving the ridge structure. 3) Binarisation and Thinning: Binarisation is the process of converting the gray scale image into the binary image. A binary image is the image where only two levels are there: 0 for the ridges which are represented by black color and 1 for the valley which are represented by white color. For binarising an image we use threshold method in which a value is presetted and the pixels lower than this threshold value are represented as white and above than this value are represented by black color. Thinning is a morphological operation that successively erodes away the foreground pixels until they are one pixel wide. Thinning is only applied to binary images and also produces a binary image as output .Thinning is the final image enhancement step performed prior to minutiae extraction .Also it is an iterative process. The thinning is applied again and again until all lines are one pixel wide .The application of thinning algorithm to a fingerprint image preserves the connectivity of the ridges and the bifurcation means the features of the fingerprints do not get distorted during thinning.

Figure 6: results of thinning an image

4. CONCLUSION
The image results shown here are obtained by applying coding in MATLAB and it is applied on the FVC 2004 database images. The results of fingerprint recognition systems are affected by the quality of the fingerprint image. The image acquired by various devices are Volume 1, Issue 3, September October 2012

S-ar putea să vă placă și