ISSN 2394-3777 (Print)
ISSN 2394-3785 (Online)
International Journal of Advanced Research Trends in Engineering and Technology (IJARTET)
Vol. II, Special Issue XIX, March 2015
FINGERPRINT IDENTIFICATION METHOD FOR SECURE
APPLICATION
RADHIKA.K.N
PG scholar
Vivekakanandha Institute of Engg. & Tech.,
for Women, Tiruchengode
ABSTRACT: Recognition of persons by means of biometric
characteristics is an important technology in the society. So
Large volumes of fingerprint are collected and stored then
used in secure applications. But it has some of the
drawbacks like required memory space is large, bad
performance, latent fingerprint images and etc. So
dictionary learning is important concept to get the
accuracy in fingerprint images. Hence adaptive orientation
model with fingerprint verification method is proposed in
efficient manner in this project. The proposed algorithm is
compared with DCT & Global based dictionary learning &
Adaptive orientation model PSNR Values. The fingerprint
verification method verifies the finger print images & find
out the nearly matches of the original images.
Keywords: fingerprint, compression, adaptive orientation,
DCT, PSNR.
I. INTRODUCTION
Fingerprint identification can’t be shared and they
intrinsically represent the individual’s bodily identity. Among
many biometric recognition technologies, fingerprint
recognition is very popular for personal identification due to
the uniqueness, universality, collectability and invariance [1].
All Rights Reserved © 2015 IJARTET
Available online at www.ijartet.com
RAMESH.M
Asst prof /ECE department
Vivekakanandha Institute of Engg. & Tech.,
for Women, Tiruchengode
A fingerprint is a procedure that maps an arbitrarily
large data item such as a computer file to a much
shorter bit string, its fingerprint, that uniquely identifies the
original data for all practical purposes just as
human fingerprints uniquely identify people for practical
purposes. This fingerprint may be used for data
duplication purposes. Fingerprints are typically used to avoid
the comparison and transmission of bulky data. For instance,
a web browser or proxy server can efficiently check whether a
remote file has been modified, by fetching only its fingerprint
and comparing it with that of the previously fetched copy
Generally compression technologies can be classed
into lossless and lossy. Lossless compression allows the exact
original images to be reconstructed from the compressed data.
By contrast, lossy data compression permits reconstruction
only of an approximation of the original data, though this
usually improves compression rates Lossless compression is
used in cases where it is important that the original and the
decompressed data be identical, or where deviations from the
original data could be deleterious. Typical examples are
executable programs, text documents, and source code. Some
image file formats, like PNG or GIF, use only lossless
compression, while others like TIFF and MNG may use either
lossless or lossy methods.
148
 ISSN 2394-3777 (Print)
ISSN 2394-3785 (Online)
Available online at www.ijartet.com
International Journal of Advanced Research Trends in Engineering and Technology (IJARTET)
Vol. II, Special Issue XIX, March 2015

Lossy compression is the class of data the dictionary, compression of a given fingerprint, quantization
encoding methods that uses inexact approximations or partial and coding and analysis of the algorithm complexity.
data discarding for representing the content that has been ARTIFACT
HISTOGRAM
encoded. Such compression techniques are used to reduce the AFFECTEDImage
Finger Print SEGMENTATION
Image
Input image denoising ANALYSIS
separation
amount of data that would otherwise be needed to store, IMAGE
handle, and/or transmit the represented content. Lossy
compression is most commonly used to
Noise Proposed Compressed
BACKGROUND
compress multimedia data audio, video, and still images, powerTHRESHOLDING
algorithm DILATE/ERODE
image
calculation CLEANING
especially in applications such as streaming media and internet
FINAL RING AREMOVED IMAGE

telephony.
Fig1. Block Diagram for Adaptive Orientation Model
The DCT-based encoder can be thought of as
compression of a stream of 8 × 8 small block of images. This
transform has been adopted in JPEG [5]. The JPEG 2.1 Image de-noising
compression scheme has many advantages such as simplicity, Image de-noising is used to improve the image
universality and availability. However, it has a bad quality by removing noise .The configuration of parallel
performance at low bit-rates mainly allows Because of the ridges and furrows with well-defined frequency and
underlying block-based DCT scheme. For this reason, as early orientation in a fingerprint image provide useful
as 1995, the JPEG-committee began to develop a new wavelet- information which helps in removing undesired noise.
based compression standard for still images, namely JPEG
2000 [6], [7].
2.2 Image separation
The DWT-based algorithms include three steps: a
Image separation is a more complicated case of image
DWT computation of the normalized image, quantization of
de-noising where more than one image is to be recovered from
the DWT coefficients and lossless coding of the quantized
a single observation. It show the effects of adaptive
coefficients. The detail can be found in [8].
dictionaries on separation of complex texture patterns from
The above algorithms are for general image
natural images.
compression. Targeted at fingerprint images, there are special
All the related studies have demonstrated the
compression algorithms. The most common is Wavelet Scalar
advantages that adaptive dictionary learning can have on the
Quantization (WSQ). It became the FBI standard for the
separation of complex texture patterns from natural images.
compression of500 dpi fingerprint images [8].
The minimization of image separation starts with extracting
and rearranging all the patches of co efficient x. The patches
II. PROPOSED METHOD
are then processed by K-SVD, which updates and estimates
sparse coefficients.
In this section, we give the details about how to use
adaptive orientation model to find the PSNR values and also 2.3 Compressed image
compress fingerprint images. The part includes construction of In compressed image using lossless compression
allows the exact original images to be reconstructed from the

149
All Rights Reserved © 2015 IJARTET

International Journal of Advanced Research Trends in Engineering and Technology (IJARTET)
Vol. II, Special Issue XIX, March 2015
compressed data. No information are loss.The
proposed method also requires the knowledge about noise
power.
Furthermore, incorporating the noise power in solving
has an important advantage in the dictionary update stage. It
ensures that the dictionary does not learn the existing noise in
the patches. Consequently, the estimated image using this
dictionary would become cleaner, which will later refine the
dictionary atoms in the next iteration. This progressive de-
noising loop is repeated until a clean image is achieved.
III. ADAPTIVE ORIENTATION MODEL
FOR DICTIONARY LEARNING
3.1 Adaptive orientation model
The adaptive orientation model is proposed for
dictionary learning process. Such dictionary learning, when
embedded into the source separation process, improves both
separation and dictionary learning quality. In other words, in
the first few iterations of Algorithm 1, each source includes
portions of other sources. However, the dictionaries gradually
learn the dominant components and reject the weak portions
caused by other sources. Using these dictionaries, the
estimated sources, which are used for dictionary learning in the
next iteration, will have fewer amounts of irrelevant
components.
Our algorithm focuses on estimating the blur kernel
and then employs a standard non-blind de-convolution
algorithm to recover the latent image. The success of our
algorithm hinges on the effectiveness and accuracy of kernel
estimation. That is, if the accuracy of the estimated kernel is
improved at each iteration.
3.2 Algorithm
 Input fingerprint image and its convert to gray scale
term.
All Rights Reserved © 2015 IJARTET
ISSN 2394-3777 (Print)
ISSN 2394-3785 (Online)
Available online at www.ijartet.com
 Scale or weight term is adaptively chosen according
to the background noise level.
 Weight term should take larger values in regions with
much structured noise and be relatively small in
fingerprint regions.
 Segmentation is done which is used to find out the
directional term.
 Orientation field estimation.
 PSNR Measurement.
IV. FINGERPRINT VERIFICATION
Fingerprint Enhancement is used to improve the
image quality by removing noise. The configuration of parallel
ridges and furrows with well-defined frequency and
orientation in a fingerprint image provide useful information
which helps in removing undesired noise.
It is the comparison of a claimant fingerprint against
an enrollee fingerprint, where the intention is that the claimant
fingerprint matches the enrollee fingerprint. To prepare for
verification, a person initially enrolls his or her fingerprint into
the verification system. A representation of that fingerprint is
stored in some compressed format along with the person's
name or other identity. Subsequently, each access is
authenticated by the person identifying him or herself, then
applying the fingerprint to the system such that the identity can
be verified. Verification is also termed, one-to-one matching.
The region mask is obtained by classifying each block
in the fingerprint image into a recoverable or unrecoverable
block and If the percentage of recoverable regions is smaller
than a threshold, the fingerprint image is rejected. An
accepted image is then passed through the filtering stage.
Two types of minutiae viz., Ridge bifurcation and Ridge
ending are considered. The minutiae are extracted based on
the number of pixels. The redundant minutiae are reduced by
150
 ISSN 2394-3777 (Print)
ISSN 2394-3785 (Online)
Available online at www.ijartet.com
International Journal of Advanced Research Trends in Engineering and Technology (IJARTET)
Vol. II, Special Issue XIX, March 2015

values was not good. In DCT PSNR values are 29.0714db and
i. All the minutiae points adjacent to each global based PSNR values are 29.5687db. Compare both our
other and within a pre-specified window is algorithm are better. It will find better PSNR values and also
removed. find clean fingerprint image.
ii. All the minutiae points near the border and
within a certain fixed distance from it are VI. SIMULATION RESULTS
removed.
iii. If two ridge endings are encountered close to
each other and if no ridges pass through
them, and then they are reconnected.

Finally the miniature is found and the fingerprint is

verified.The similarity images also found in the false
miniature process.

V. COMPARISION FOR EXISTING Fig. 2. Clean Image by adaptive dictionary and PSNR
Values

Clean Image Psnr Values

By Dictionary

DCT 29.0714 db

GLOBAL 29.5687 db
BASED Fig.3. Dictionary Trained on Patches From Compressed
Image,

VII. CONCLUSION
A new compression algorithm adapted to fingerprint

TABLE .1 images is introduced.In our proposed algorithms, there with

existing more sophisticated algorithms. We consider the effect
In existing method, use two different algorithm DCT of DCT and global based algorithms different clean images
and global based. In DCT are used lossy compression, bad by dictionary on fingerprint compression and also find
performance because of the underlying block-based DCT different PSNR values, compare these existing algorithm both
scheme and global based also have same problem, both PSNR have some drawbacks. Using our proposed algorithm gives
151
All Rights Reserved © 2015 IJARTET
 ISSN 2394-3777 (Print)
ISSN 2394-3785 (Online)
Available online at www.ijartet.com
International Journal of Advanced Research Trends in Engineering and Technology (IJARTET)
Vol. II, Special Issue XIX, March 2015

better PSNR values, clean fingerprint [9] A. Said and W. A. Pearlman, “A new, fast, and
verification and gives efficient output. efficient image codec based on set partitioning in
It mainly used to all secure purpose for all in our hierarchical trees,” IEEE Trans.Circuits Syst. Video
world. Fingerprint will give unique identification.Using this Technol., vol. 6, no. 3, pp. 243–250, Jun. 1996.
method to avoid overlapping problem also. [10] A.Skodras, C. Christopoulos, and T. Ebrahimi, “The
JPEG 2000 still image compression standard,” IEEE
REFERENCES Signal Process. Mag., vol. 11, no. 5, pp. 36–58, Sep.
[1] N. Ahmed, T. Natarajan, and K. R. Rao, “Discrete 2001.
cosine transform,” IEEE Trans. Comput., vol. C-23, [11] R. Sudhakar, R. Karthiga, and S.Jayaraman,
no. 1, pp. 90–93, Jan. 1974. “Fingerprint compression using contourlet transform
[2] C. M. Brislawn, J. N. Bradley, R. J. Onyshczak, and T. with modified SPIHT algorithm,” IJECEIranian J.
Hopper, “FBI compression standard for digitized Electr. Comput. Eng., vol. 5, no. 1, pp. 3–10, 2005.
fingerprint images,” Proc. SPIE,vol. 2847, pp. 344–
[12] Y. Y. Zhou, T. D. Guo, and M. Wu, “Fingerprint
355, Aug. 1996.
image compression algorithm based on matrix
[3] C. S. Burrus, R. A. Gopinath, and H. Guo,
optimization,” in Proc. 6th Int. Conf. Digital Content,
Introduction to Wavelets and Wavelet Transforms: A
Multimedia Technol. Appl., 2010, pp. 14–19.
Primer. Upper Saddle River, NJ, USA: Prentice-Hall,
1998.
[4] Guangqi Shao, Yanping Wu, Yong A, Xiao Liu, and
Tiande Guo, “Fingerprint Compression Based on
Sparse Representation”, IEEE Transactions on image
processing, vol. 23, NO. 2, Feb. 2014.
[5] D. Maltoni, D. Miao, A. K. Jain, and S. Prabhakar,
Handbook of Fingerprint Recognition, 2nd ed.
London, U.K.: Springer-Verlag, 2009.
[6] W. Marcellin, M. J. Gormish, A. Bilgin, and M. P.
Boliek, “An overview of JPEG-2000,” in Proc. IEEE
Data Compress. Conf., Mar. 2000, pp. 523–541.
[7] W. Pennebaker and J. Mitchell, JPEG—Still Image
Compression Standard. New York, NY, USA: Van
Nostrand Reinhold, 1993.
[8] S. Prabhakar, “Fingerprint classification and matching
using a filterbank,” Ph.D. dissertation, Dept. Comput.
Sci., Eng., Michigan State Univ., East Lansing, MI,
USA, 2001.

152
All Rights Reserved © 2015 IJARTET