Fingerprint Scholar (ECE) NITTTR , Chandigarh [email protected] Rohini ME

Fingerprint Enhancement Using Frequency Domain FilterAnavil MishraME Scholar (ECE) NITTTR , [email protected]                                                                                                       RohiniME Scholar (ECE) NITTTR , [email protected]  In the present paper we will discuss aboutfingerprint image enhancement for fingerprint recognition. Many experiments have beendone in past for fingerprint image enhancement. Many new procedures &advances have developed gradually which has improved the poor qualityfingerprint images in a remarkable way. Therefore, in the given paper, anassessment of these procedures & advances are done with its rationalepoints to validate the results. Based on the given procedure, a consolidatedsolution for fingerprint identification is evolved for affirmation.

For the plan of action some methods at the time of coding &calculations are initiated to improve & upgrade the working of thisfingerprint recognition system. The results are proposed only after conductingmultiple experiments on variety of fingerprint images Keywords: Biometrics, Image processing, Fast Fourier transform, GaussianBand pass Filter, signalprocessing, image analysis, feature extraction, fingerprint image enhancement,minutia matching.1.    IntroductionBIOMETRICS is one of the most reliable solutions to determine identityof an individual by identifying their physiological or behavioralcharacteristics. Some of the commonly used physiological characteristics andbehavioral for biometric recognition are face, hand & fingerprints, iris,retina, DNA, signature, voice etc. Automated fingerprint identification system(AFIS) is the most commonly used technique implemented across the globe. It hastherefore, now become most important in security field because of which it hasinterested many researchers to continue doing their research on it.

Best services for writing your paper according to Trustpilot

Premium Partner
From $18.00 per page
4,8 / 5
Writers Experience
Recommended Service
From $13.90 per page
4,6 / 5
Writers Experience
From $20.00 per page
4,5 / 5
Writers Experience
* All Partners were chosen among 50+ writing services by our Customer Satisfaction Team

Every humanbeing have different patterns of crest (ridges) & troughs (valleys). Theimage which is shown in Fig.1 is showing the surface of the fingerprint whichis taken where the dark lines are showing the ridges and the light lines in thefingerprint image is showing the valley.   Fig. 1.  Ridge and Valley Structure 1  Mainly, three standards about fingerprint arementioned below:  1.1Unalterable: Fingerprint patterns are permanent and never alter during thegrowth of human body.

1.2.Unique: Every human being has unique fingerprint ridges & arenever identical in two different people. 1.3.

Grading: Grading of fingerprint patterns is limited. This makes thelogical grading of fingerprint pattern easier. Images of fingerprint which areobtained from people for identification purposes are not often clear andsometimes quality is also not perfect. Different factors like scars, sometimes moisturewhich is present in scanner and there are also many more reasons which areresponsible for affecting the quality of an image which can lead to failure ormaybe there are chances of false extraction of minutiae. Hence, it has become necessary toadapt techniques for image enhancement to reduce noise present in it and alsoto improve statement of ridges against valleys.

This paper contains, variousfingerprint image enhancement techniques and these are discussed. The outcome resultof these techniques are shown in the given Table 1 and results of thesetechniques are also discussed in Section 5. This further propose the fingerprintimage enhancement in two different domains that are spatial domain &frequency domain which  are discussed. Clarityof fingerprint image and contrast of different fingerprint images are improvedby using technique that is the spatial domain technique and then GaussianBandpass filter is used which is tuned to afixed range of frequencies to transform them to the different fourier image.  Resultants are experimented on originalsamples of fingerprints by using filters to further analyze the frequencydomain which is generated directly. 2.

     PROPOSED WORK& BACKGROUNDIn this section, to improve theclarity of ridges & valleys of the fingerprint impression some calculations& methods are initiated that have been initiated to improve the pictureclarity of ridges and valleys. In this Frequency Domain Filter is used whichallows the fingerprint images to combine with the large size filtersefficiently. For the improvement of imageGabor filters have been used as initiated in 7, 8, 9, 10. To use Gaborfilters,  four different parameters areused that are local ridge orientation, local ridge frequency and the standarddeviation of the Gaussian envelop along the x-axis and y-axis (?x and ?y) must beselected specifically to avoid false ridges & valleys.

For otherfingerprint enhancement techniques the methodology has been explained inSection 3 & the results of the methodology implemented is given in Section5.  3.     FINGERPRINT IMAGEENHANCEMENT TECHNIQUES The following techniques have been used to analyse& enhance the fingerprint images: 3.

1.  Short Time FourierTransform (STFT) Analysis:In 2 3, as anew technique STFT analysis has been introduced for fingerprint magnification.Fig.

2 below shows the audit of the initiated methodology.           Originalimage                                                                                                                           Enhanced Image                    Fig. 2. Initiated methodology in 2 Basedon the contextual filtering a new fingerprint image enhancement algorithm inthe Fourier domain is presented. In Fig. 1, input image is divided intoprojected windows during STFT analysis. It is presumed that the image withinthis small window is soaked and can be modelled as a surface wave. The roughcalculations of ridge frequency and ridge orientation are then acquired.

  For the computation of angular coherence theenergy map yielded from STFT analysis is used. This coherence image is used toadapt to the angular bandwidth. Then, the resultant contextual information isused to filter each window on the Fourier domain.     table  Theabove calculated table can be summarised as under:Ithas two stages. The first stage is composed of STFT analysis and second stageis composed of contextual filtering. The STFT stage furnishes the three images-ridge orientation, ridge frequency and the block energy. These are furtherutilised to calculate the region mask.

This is almost identical to thetechnique which is initiated in 4. Hence, all the intrinsic images which arerequired to perform contextual filtering are produced out of the analysis phase. 3.2.   Pyramid-based FilteringThis technique uses thefamiliar methods for improvement and for drawing particulars as initiated in5. Two types of symmetries namely parabolic and linear are used to prototypeas an example & draw the confined structure in a fingerprint image &these symmetries can be roughly calculated by separate filtering of theorientation tensor.

In the improvement stage, a Laplacian-like image pyramid isused to fragment the original fingerprint image into sub bands correlating todifferent spatial scales.The frequency directions originate from the frequencystructures where surround smoothing is done on the pyramid at different levels.In Fig.

4, the reduction (I, k) and expansion (I, k) function are defined tocreate the Gaussian & Laplacian pyramid. The function is shown below inFig. 4.                                                         a) Pyramid Decomposition PD  a)       Gaussian-Like Laplacian-Like g1 = reduce (fp, ko); g2 = reduce (g1,k); g3 = reduce (g2,k); g4 = reduce (g3,k); l1 = g1- expand(g2,k); l2 = g2-expand(g3,k); l3 = g3- expand(g4,k)   b)       Reconstruction R fp = expand (……ko);                              expand (l3,k) +l2  Fig. 4.

 Pyramid building processin 5 Laterto detect the accurate position & direction of the minutia the parabolicsymmetry is added to the local fingerprint model. The parabolic symmetry isfurther reduced if the linear symmetry is high whereas it is maintained viceversa. All these steps are followed in compliance with extracting the finerdetails.3.3.  Curved Gabor FiltersTheGabor Filters play a crucial role in the improvement of different types ofimages & in the process of drawing Gabor features. In 6, the curved Gaborfilters were introduced purposely to upgrade the curved structures in noisyimages.

The choice of filter parameters increases the smoothing power of themagnified images without fabrication. In severaloperations of image processing and pattern realization, Gabor functions 8, inthe form of Gabor filters 9 and Gabor wavelets 10 were enforced for variousfunctions In6, it is observed that curved Gabor filter is applied to the curved ridge& valley structure to enhance the poor quality fingerprint images.Initially, two orientation field estimation methods were combined to obtainmore sturdy estimations for fingerprint images which are extremely noisy innature. CurvedGabor filters allow the use of large curved regions without creating falsefeatures. It is therefore a testified & proven fact that the curved Gabor Filters drawbetter results in comparison to the traditional Gabor Filters.   Fig. 5. (Left): OriginalFingerprint Image, (Middle): Image drawn after the use of Traditional Gaborfilter, (Right): Image drawn after the use of Curved Gabor filter.

 3.4.  FastFourier Transform and Gaussian Bandpass Filters Have been used to magnifythe fingerprint images. 3.4.1.

 The subdivided image is further divided intonon converging blocks of size W x W as presented in 7, The total time takento calculate FFT is much faster when the block sizes are power of two8.Thus,16 X 16 is suggested to use for computation of value.3.4.2.  Further for every block, 2D Discrete FastFourier Transform is suggested to apply based on 1. The value F(0,0) of theDFT is called the DC coefficient.

When u=v=0 is substituted in the definitiongiven in 1, then F(0,0) = ? M-1 ? M-1  f(x,y) exp(0 ) = ? M-1 ? M-1  f(x,y)                                                                                         (2)                                                                                                                        X=0   y=0                                   x=0   y=0 The Fouriertransform is complex. If the real components are represented by R(u,v) andImaginary components are represented by I (u,v) then the Fourier range will bedefined as:  =           To envision, it is appropriate to havezero frequency element DC in the center of the matrix 8. In comparison toother values DC coefficient is larger.

Fourier transform Fingerprint image isupgraded by a log transformation as in 8 & is shown in fig. 6   Fig. 6. Spectrum of the image block 3.

4.3.  In Fig. 7 extent of frequencies in Bandpassfilter is shown.

The cut-off frequency values for lower & higher limit aretaken as 130 and 225 respectively for sharpness. Multiply the FFT spectrum ofthe block by the filter H (u,v)       G = H ? F3.4.

4.Counter FFT transform is used to recover thefiltered image. 3.4.5.

On repeating the above mentioned process theenhanced image is obtained after the initial calculations are applied as shownin Fig. 8   Fig. 7. Imageobtained after using Gaussian Bandpass Filter   Fig.

8.Magnified image after using Gaussian Bandpass Filter  4.    RESULT:4.

1.  FORSTFT , PYRAMID BASED FILTERING  &CURVED GABOR FILTERING   In the given Table 1 the outcome of validityof techniques used for fingerprint image magnification are given, based on theEER’s percentage from the review papers mentioned above. TABLE 1. RESULTS OFFINGERPRINT IMAGE ENHANCEMENT TECHNIQUES                                                                                                   Author Techniques Result/Accuracy Bhowmik,etal.

1 DFT and Histogram Equilization Quality of fingerprint image greatly increased and more accurate minutiae extraction Chikkerur, Cartwright, Govindaraju 2 Short Time Fourier Transform (STFT) Analysis % of EER (FVC2004)          DB1 = 19.1          DB2 = 11.9          DBE = 7.6          DB4 = 10.9   Table 1 (Continued)  Fronthaler Kollreider, Bigun 5 Pyramid- based filtering % of EER (FVC2004)          DB1 = 12.0          DB2 = 8.

2          DBE = 7.6          DB4 = 7.0 Gottschlich 7 Curved Gabour Filters % of EER (FVC2004)          DB1 = 9.7          DB2 = 6.3          DBE = 5.1          DB4 = 6.5   EER = Equal ErrorRate, FAR = False Acceptance Rate, FRR = False Rejection Rate (FROM THIS TABLEREMOVE THE FOLLOWING HEADINGS AS I HAVE NOT INCLUDED THEM IN THIS REVIEW PAPER– 1,5,7)EERvalue given in the above table shows that the segments of both FAR & FRRare equal. The lower the EER value, higher the authenticity of the biometricsystem.

The functioning of curved Gabor filters 6 depends on the nature of the orientation fields (OF) and RF approximation. 4.2.

  RESULTFOR FAST FOURIER TRANSFORM & GAUSSIAN BANDPASS FILTERS WILL CONCLUDE THEANSWER  Assessmentswere conducted on as many as 138 Fingerprint images gathered from 46 differentpeople from all class & categories using an Optical Fingerprint Reader. Thesize of each image is 480 x 320. Highpass and Bandpass frequency domain filterswere used to affirm the best technique used in the process of fingerprintimprovement.Thefollowing guidelines & framework are investigate the methodology discussedabove. 4.

2.1. PSNR: PeakSignal-to-noise ratio ThePeak Signal to Noise Ratio (PSNR) produce a degreeofpeak error in decibels between two images. This ratio is often used as aquality measurement between the original image and a compressed orreconstructed image 9. The technique with a higher PSNR produce betterresults. PSNR is expressed as: PSNR = 10log10                                                                                                                                            HereR is the ultimate desirable pixel value in the input image. The value of R istaken as 255 for an 8-bit unspecified total data type gray scale image. The MSEproduces the aggregate squared error between the recreated & the pioneerimage.

Here, the lower value of MSE shows lower error in the reformation of animage. The mean-squared error is given as:   =    I(I , j) – K ( I , j)2                                                                                   where,M & N are the number of rows and columns in the suggested imagessubsequently. Image I is to be investigated & image K will work as therelating image. 4.2.2.

  CalculationTime: This measure is used to guage thetotal time taken to magnify the image & can be ascertained by using cputimecommand of Matlab. Table-IIshows the conclusions drawn empirically by implementing frequency domainfilters on the sample Fingerprint images using Matlab.  TABLE IIEXPERIMENTALRESULTS      Enhancement Technique Peak Signal to Noise ratio Computation Time Gaussian High pass Filter 7.4657db 0.8892 secs Gaussian High pass Filter 26.

4865db 0.8644 secs   5.     CONCLUSION:  The realization of afingerprint feature derivation & coordination calculations largely rely onthe quality of fingerprint images used for desired results. There have beenconstant efforts made in enhancing the quality of fingerprint images as shownin different studies. The oriented diffusion filter & curved Gabor filterprocess give preferred results. In the latterhalf of the paper the other methods & techniques used for fingerprintenhancement such as, Gaussian Bandpass filter is used in the frequency domain.

Here, with the use of PSNR the filtered image is capable of reducing the noise& smoothening the holes & small breaks in a ridge in less time framerequired for calculation. The Fourier spectrum can be rotated undeviatinglywith the developed Bandpass Filter tuned to a limited range of cut off frequencies.This work has concentrated on the circularly proportionate filters that arestated as functions of the distance from the center of the filters. Thefuturistic opportunity of this task can be to investigate more frequency domainfilters to accomplish improved outcomes.

 6.    REFERENCES:  1 Davide Maltoni, Dario Maio, Anil K. Jain,Salil Prabhakar, HandBook of Fingerprint Recognition, Second Edition,Springer-Verlag (2009).   2 Shaikh Mohammedsayeemuddin, Sima KGonsai and Dharmesh Vandra, “Efficient fingerprint image enhancement algorithmbased on Gabor filter”, International Journal of Research in Engineering andTechnology, Vol.3, pp. 809-813, 2014.3Yuan Mei, Bo Zhao, Yu Zhou, Songcan Chen, “Orthogonal curvedline Gabor filterfor fast fingerprint enhancement”, Electronics Letters, Biometrics Compendium,IEEE, Vol.

50, Issue 3, pp.175 – 177, 2014.4Miao-li Wen, Yan Liang, Quan Pan, Hong-cai Zhang, “A Gabor filter basedfingerprint enhancement algorithm in Wavelet domain”, IEEE International Symposiumon Communications and Information Technology, Vol.

2, pp.1468- 1471,20055L. Hong, Y. Wan, and A. Jain, “Fingerprint image enhancement: Algorithm andPerformance Evaluation”, IEEE Transactions on Pattern Analysis and MachineIntelligence, vol. 20, pp.

777-789, 19986Miao-li Wen, Yan Liang, Quan Pan, and Hong-cai Zhang, “A Gabor filter basedfingerprint enhancement algorithm in Wavelet domain”, In proceeding of IEEEInternational Symposium on Communications and Information Technology, Vol2,pp.1421-1424, 2005 7S. Chikkerur, A.N. Cartwright and V. Govindaraju, “Fingerprint EnhancementUsing STFT Analysis”, Pattern Recognition, Vol.

40, No. 1, pp. 198-211, 2007.(Journal) 8S.

Chikkerur, A.N. Cartwright and V. Govindaraju, “Fingerprint imageenhancement using STFT analysis,”Third International Conference on Advances inPattern Recognition, ICAPR 2005, Bath, UK, Proceedings, Part II, pp. 20–29,2005. (Conference Proceedings) 9L.

Hong, Y. Wan and A. Jain, “Fingerprint Image Enhancement: Algorithm andPerformance Evaluation”, IEEE Trans. On Pattern Analysis and MachineIntelligence, Vol. 20, No.

8,pp. 777-789, 1998. (IEEE Transactions) 10 H. Fronthaler, K.Kollreider and J.

Bigun, “Local Features for Enhancement and Minutiae Extraction in Fingerprints”, IEEE Trans. ImageProcessing, Vol. 17, No.

3, pp. 354-363, 2008. (IEEE Transactions) 11C. Gottschlich, “Curved Gabor Filters for Fingerprint Image Enhancement”, IEEETrans.

On Image Processing, Vol. 21, No. 4, pp.

2220-2227, April 2012. (IEEETransactions)12D. Gabor, “Theory of Communication. Part 3: Frequency Compression andExpansion”, Journal of the Institution of Electrical Engineers, Vol.

93, No.26, pp. 429-441, 1946. (Journal)13D.

Gabor, “Information Theory in Electron Microscopy”, LaboratoryInvestigation, Vol.14, No. 6, pp. 801-807, 1965. (Article)14T.

S. Lee, “Image Representation Using 2D Gabor Wavelets”, IEEE Trans. onPattern Analysis and Machine Intelligence, Vol. 18, No. 10, pp. 959-971, 1996.(IEEE Transactions)   15 Kumud Arora, Poonam Garg, “AQuantitative Survey of various Fingerprint Enhancement techniques”,International Journal of Computer Applications, Volume 28, pp.24-29, 2011