International Journal of Scientific Research in Engineering and Management (IJSREM)

Volume: 05 Issue: 09 | Sept - 2021 ISSN: 2582-3930

Finger Print Image Recognition-A Review

Mrs. S.Subhashini , Assistant Professor

Dept.Of.Commerce with computer Science, Mannar Thirumalai Naicker college

Abstract
Fingerprint recognition is one of the most well- To avoid the extraction of false minutiae, there are many
known and publicized biometrics. Because of finger print steps which should be followed such as preprocessing and
uniqueness and consistency over time, fingerprints have enhancement. One of these steps is fingerprint
been used for identification for more than a century. Though segmentation.
Many systems with minutiae extractor and minutiae matcher Fingerprint segmentation is the process by which
are available for minutiae extraction and matching ,it is very
the foreground is separated from the image background. The
difficult to mark all the minutiae accurately as well as
result of fingerprint segmentation is a fingerprint image in
rejecting false minutiae because of the presence of noise in
fingerprints. . In practise, it is usually difficult to take a which the background is removed .
good quality fingerprint image, as these may be degraded The aim for fingerprint segmentation is to separate the
and corrupted with noise due to many factors including foreground from the background. Due to the nature and the
variations in skin and impression conditions. This poor quality of fingerprint image, the finger print
degradation can result in a significant number of spurious segmentation becomes an important and challenging task.
minutiae being created and genuine minutiae being ignored. Fingerprint segmentation is an important step in the
Image enhancement techniques are employed prior to automatic fingerprint recognition systems because it
minutiae extraction to obtain a more reliable estimate of
improves the fingerprint images so that features can be
minutiae locations. In recent years, some new methods
extracted from these images by the automatic fingerprint
have been introduced to the finger print image recognition
system to recognize finger print images in order to get recognition systems.
better results . In this paper, Finger Print Image recognition
and a methods used for minutiae extraction and matching 2. FINGER PRINT IMAGE AND MINUTIAE
, pros and cons of this method are discussed. A fingerprint image is a pattern which consists of
Keywords: Image Recognition , finger print, Minutiae, two regions, foreground and background. The foreground
pixel features , Crossing number, Biometrics contains all important information needed in the automatic
fingerprint recognition systems. In general, the fingerprint
1. INTRODUCTION of an individual is unique and is formed from an impression
of the pattern of ridges on a finger. A ridge can be defined
Biometric identification using fingerprints matching as a single curved segment, and a valley is the region
have been used for last many years and are most widely between two adjacent ridge. The term minutiae refer to the
used form of biometric identification. Despite this local discontinuities in the ridge flow pattern and provide
widespread use of fingerprints, there has been little the features that can be used for biometric identification.
statistical work done on the uniqueness of fingerprint The characteristics such as orientation and location of
minutiae. More specifically, identifying the characteristics minutiae are usually taken into account when performing
of minutiae in fingerprint is a critical task and is thus an fingerprint matching. A typical example of fingerprint,
active area of research. In Biometric applications , Finger ridges, valleys and minutia is shown in Fig 1 below.
print images are used to identify an individual and verify
their identity.A fingerprint image is a pattern which consists
of two regions, foreground and background. The foreground
contains all important information needed in the automatic
fingerprint recognition
systems. However, the background is a noisy region that
contributes to the extraction of false minutiae in the system. FIG 1

© 2021, IJSREM | www.ijsrem.com | Page 1

 International Journal of Scientific Research in Engineering and Management (IJSREM)
Volume: 05 Issue: 09 | Sept - 2021 ISSN: 2582-3930

previously stored in the system database. Therefore the

3. FINGER PRINT RECOGNITION system returns either a degree of similarity in case of
Fingerprint recognition is the process of comparing a identification or a binary decision in case of verification.
fingerprint against another fingerprint to determine if 4. Minutiae Marking and Extraction :
the impressions are from the same finger. It is very
difficult to mark all the minutiae accurately as well as To mark and extract the minutia, the key
steps are as follows:
rejecting false minutiae because of the presence of
noise in fingerprints. Fingerprint Ridge Thinning: Ridge Thinning is done to
Fingerprint recognition systems consist of the eliminate the redundant pixels of ridges till the ridges are
following parts : just one pixel wide. This is done using the morphological
1.Sensing or Image acquisition thinning function. bwmorph(binaryImage,‘thin’,Inf). The
2.Pre-processing thinned image is then filtered, again using three
morphological functions to remove some H breaks, isolated
3.Feature or minutiae extraction
points and spikes using the bwmorph(binaryImage,
4.Matching
‘hbreak’, k); bwmorph(binaryImage, ‘clean', k) and
bwmorph(binaryImage, ‘spur', k).
Minutiae Marking: After the fingerprint ridge thinning, the
next step is to mark minutia points. This can be done in
three ways: 1) for each 3x3 window, if the central pixel is 1
and has exactly 3 one-value neighbors, then the central pixel
is a ridge branch (Figure 7a). If the central pixel is 1 and has
Figure 2. Fingerprint Recognition System
only 1 one-value neighbor, then the central pixel is a ridge
ending Suppose both the uppermost pixel with value 1 and
Sensing or Image Acquisition the rightmost pixel with value 1 have another neighbor
The acquisition of a fingerprint images was outside the 3x3 window, so the two pixels will also be
accomplished by using off-line sensing or live-scan. Off- marked as branches.
line sensing is defined as ink-technique. Live-scan False Minutiae Removal: This stage focuses on removing
scanners become presently more frequent, because of its any false minutia, e. g., any false ridge breaks due to noise.
simplicity in usage. There is no need for ink. The digital For this purpose, the average inter-ridge width D is
estimated at this stage. The average inter-ridge width refers
image is directly acquired by pressing against the surface
to the average distance between two neighboring ridges and
of the scanner. for each row it is given by:
Pre-processing
To simplify the task of minutiae extraction and Inter ridge distance = sum all pixels with value 1 /
make it more easy and reliable, some preprocessing row length
techniques are applied to the raw input image. Enhancement
and segmentation of the fingerprint are the most commonly Finally an averaged value over all rows gives D. Then, all
the thinned ridges in the fingerprint image are labeled with
methods performed in the preprocessing step.
a unique ID for further operation using morphological
Feature extraction operation BWLABEL. Seven types of false minutia are
After preprocessing step, the segmented and specified in this paper as described in Table 1. For better
enhanced fingerprint is further processed to identify the results, this paper considers all the seven cases unlike the
main and distinctive minutiae. Most of the minutiae other similar works done. The minutiae and after removing
extraction methods necessitate the fingerprint gray-scale false minutiae for the above two finger print images is
image to be transformed into a binary image. The acquired shown in Figure 3.
binary image is forwarded to a thinning stage to reduce the
thickness of the ridge to one pixel ridge. Afterwards, the TYPE DESCRIPTION
m1 A spike piercing into a valley.
minutiae are simply detected by a simple image scan.
Matching m2 A spike falsely connects two
Algorithms that extract important and efficient Ridges.
minutiae, will improve the performance of the fingerprint m3 Two near bifurcations located in
matching techniques. The features extracted of the input The same ridge
image are compared to one or more template that was m4 The two ridge broken points have nearly

© 2021, IJSREM | www.ijsrem.com | Page 2

 International Journal of Scientific Research in Engineering and Management (IJSREM)
Volume: 05 Issue: 09 | Sept - 2021 ISSN: 2582-3930

the same orientation and a short distance. two fingerprints matches or not. As the % match obtained is
m5 This is alike the m4 case with the less than the threshold so the finger prints do not match
Exception that one part ofvthe broken which is correct.
ridge is so short that another termination
is generated. 5. THINNING
m6 This case extends the m4 case Thinning is a process of extracting a skeleton from
But with the extra property that a third an object in a digital image. A skeleton of an image can be
ridge is found in the middle of the two
parts of the broken roidge. thought of as a one-pixel thick line through the middle of
m7 Only one short ridge found in the an object which preserves the topology of that object.
threshold window Thinning is a fundamental preprocessing step in many
image processing and pattern recognition algorithms .
Thinned images (skeletons) are easier to process and they
reduce processing time for the subsequent operations.
Many thinning algorithms have been developed in the past
three decades. Two major approaches of thinning digital
patterns can be categorized into iterative boundary removal
algorithms and non-iterative distance transformation
algorithms .
Iterative boundary removal algorithms delete pixels on
the boundary of a pattern repeatedly until only unit pixel-
width thinned image remains. Non-iterative distance
transformation algorithms are not appropriate for general
applications since they are not robust, especially for patterns
Fig 3
with highly variable stroke directions and thicknesses.
MINUTIAE MATCHING Thinning based on iterative boundary removal can be
divided into sequential and parallel algorithms .
After minutiae extraction for the two fingerprint images, the
next step is to match the Minutiae. For this purpose, an
elastic string based iterative ridge alignment algorithm is
used. In this algorithm first the minutiae’s of two
fingerprints are aligned and then the percentage of the
matched minutia pairs is computed. In the alignment stage,
the two fingerprint images to be matched are taken and any
one minutia from each image is chosen. Then the similarity
of the two ridges associated with the two referenced minutia
points are calculated using the standard cross-correlation
formula. If the similarity is larger than a threshold (0.8
chosen in current paper, as maximum is 1), each set of the In sequential algorithms, the pixels are examined for
minutia is transformed to a new coordination system whose deletion in a fixed sequence in each iteration, and the
origin is at the referenced point and x-axis is coincident deletion of pixel p in the nth iteration depends on all
with the direction of the referenced point. To better operations
understand, let M ( , , ) be the reference minutia (say
performed so far, i.e. on the results of (nth 1)th iteration;
from I1). Then for each fingerprint, all other minutiae ( ,
as well as on the current pixel in the n iteration. In a
, ) is translated and rotated with respect to the M . The parallel algorithm,the deletion of pixels in the nth iteration
above implementation was an effort to understand how depends only in the results of the nth iteration; therefore, all
Fingerprint Recognition is used as a form of biometric to pixels are examined independently in the parallel manner in
recognize identities of human beings. It includes all the each iteration The behavior of a thinning algorithm is
stages from minutiae extraction from fingerprints to determined by its structuring element. Structuring elements
minutiae matching which generates a match score. Various are policies which define the situations at which foreground
standard techniques are used in the intermediate stages of pixels will be set to background and hence deleted. Thinning
processing.It is implemented for two different fingerprints is used in but not limited to applications that process
shown above in this paper and the percentage of matching handwritten and printed characters, fingerprints and palm
of minutia is computed. For the two finger prints the prints, chromosomes and biological cell structures, and
percentage match obtained is 34.6154%, which when circuit diagrams.
matched with a threshold limit (say 90%) can say that the

© 2021, IJSREM | www.ijsrem.com | Page 3

 International Journal of Scientific Research in Engineering and Management (IJSREM)
Volume: 05 Issue: 09 | Sept - 2021 ISSN: 2582-3930

Generally, fingerprint recognition systems work by thinning algorithms are not suitable for all applications.
matching minutiae extracted from probe data, to reference Appropriate algorithm for particular application must be
minutiae and it consists of the following stages: fingerprint chosen.
acquisition, image pre-processing (fingerprint segmentation,
enhancement, and orientation field estimation), fingerprint REFERENCES
classification, minutiae detection and matching . Fingerprint
1. A.K. Jain, Fundamentals of Digital Image
thinning is an important image enhancement processing step Processing, Prentice-Hall, Englewood Cliffs, NJ,
in an Automatic Fingerprint Identification System (AFIS). It 1989.
plays an equally significant role with fingerprint
classification and enhancement in practical AFIS. It can 2. Afsar, F.A., Arif, M. and Hussain, M. 2004.
Fingerprint Identification and Verification System
significantly improve the recognition performance of an
using Minutiae Matching. In proceedings of the
AFIS. Binary image thinning has been studied extensively National Conference on Emerging Technologies. 141-
in literature. While some researchers have developed 146.
sequential algorithms , the main focus is in parallel thinning 3. Akram, M., Nasir, S., Tariq, A., Zafar, I. and Khan, W. S.
algorithms, which are efficient and fast. Raju and Xu in 2008. Improved Fingerprint Image Segmentation Using
New Modified Gradient Based Technique. In proceedings
their study of parallel thinning algorithms compared Zhang-
of the 2008 Canadian Conference on Electrical and
Suen, Guo-Hall and One Pass Thinning Algorithm (OPTA) Computer Engineering. Niagara Falls, Canada, 001967 –
for character recognition. They found that Guo-Hall 001972
outperformed the two other algorithms in terms of skeleton 4. D. Maio, and D. Maltoni, “Direct gray-scale minutiae
detection in fingerprints”, IEEE Transactions Pattern Analysis
quality. While OPTA is faster than the other two algorithms,
5. and Machine Intelligence, vol. 19(1), pp. 27-40, 1997.
its skeleton quality is not as good compared to those of the
other two algorithms. Gupta and Kaur compared Zhang- 6. Robert Hastings, “Ridge Enhancement in Fingerprint Images
Using Oriented Diffusion”, IEEE Computer Society
Suen, Abdulla et al and a multipass iterative boundary
on Digital Image Computing Techniques and Applications,
removal algorithm based on. They found that the mutlipass pp. 245-252, (2007).
algorithm produced better results than Zhang-Suen and
7. Eric P. Kukula, Christine R. Blomeke, Shimon K. Modi, and
Adbulla et al with regards to connectivity and spurious Tephen J. Elliott, “Effect of Human Interaction on
branches of numerical patterns. Fingerprint Matching Performance, Image Quality, and
Minutiae Count”, International Conference on Information
Technology and Applications, pp. 771-776, (2008).
6. CONCLUSIONS 8.
Minutiae Extraction and matching heavily influences 8. Bazen, A. M., Verwaaijen, G. T. B., Gerez, S. H.,
the performance of fingerprint recognition system. In recent Veelenturf, L. P. J. and Zwaag, B. J. 2000. A Correlation-
years, some new methods have been introduced to the Based Fingerprint Verification System. In proceedings of
the ProRISC Workshop on Circuits, Systems and Signal
minutiae extraction and matching in order to get better
Processing. Veldhoven,Netherlands.
disposal results. Thinning plays a very important role in the
prepro-cessing phase of automatic fingerprint 9. Gonzalez, R. C. and Woods, Richard E. 2008. Digital
recognition/identification systems. The performance of Image Processing.
minutiae extraction relies heavily on the quality of skeletons
used. A good fingerprint thinning algorithm can depress 10. Greenberg S. , Aladjem, M., Kogan, D. and Dimitrov, I.
image noise and promote the robustness of the minutiae 2000. Fingerprint Image Enhancement using Filtering
extraction algorithm which helps improve the overall Techniques. In proceedings of the 15th International
performance of the system. Many thinning algorithms have Conference on Pattern Recognition. Barcelona, Spain,
227-- 236.
been devised and applied to a wide range of applications
including, Optical Character Recognition (OCR), biological
cell structures and fingerprint patterns.Choosing both
minutiae extraction and thinning algorithms for a particular
application is very difficult. All minutiae extraction and

© 2021, IJSREM | www.ijsrem.com | Page 4