Sie sind auf Seite 1von 4

International Journal of Excellence Innovation and Development

||Volume 1, Issue 1, Nov. 2018||Page No. 046-049||

Artificial Neural Network Based Offline


Signature Recognition System Using Local
Texture Features
Shivashankar M. Rampur
Professor, Department of Computer Science and Engg., Brahmdevdada Mane Institute of Technology, Solapur Belat Tal.North
Solapur, Distt. Solapur, Maharashtra, India

Abstract––Signature of an individual is a significant local features that classifies each input signature based on
biometric attribute of a human being which can be used artificial neural network and has many number of uses. The
to certify human identity and these attributes can have neural networks is the most outstanding way of finding
own identity like face recognition, fingerprint detection, solution of the problems that are most difficult to solve by
iris inspection and retina scanning. In this work, we are traditional computational methods. The advantage of neural
developing method which deals with the off-line network is no need to understand the solution.
signature recognition system using artificial neural
network in which the signature is captured and presented While signer is signing, there are variations in terms of pen
in the form of an image to the system. Offline signature width, additions found in strokes, exchange or qualified
recognition system is a significant biometric technique, point of strokes, scaling within the genuine signatures and
which is used to offers automated process of recognition rotation. Our system is motivated to overcome these
and verification by extracting local features that variations. Our system gives high level of accuracy.
classifies each input signature and has many number of
uses. The proposed system has local texture features and Objectives
feed forward back propagation in an artificial neural  The objective of our system is used to develop
network classifier to identify and authenticate signatures preprocessing phase which is processed on input
of individuals. Various image processing techniques are signature image. This preprocessing phase include
used to categorize and validate the signature. conversion of original image into grey scale,
conversion of grey into binary image, noise
Index Terms––Image acquisition, RGB-to-Grayscale reduction, thinning and resize.
conversion, feature extraction, artificial neural network  The objective of our system is used to develop
feature extraction phase for classifying signature. In
this phase, we are extracting texture features from
INTRODUCTION
signature which are entropy, homogeneity, contrast,
Signature is generally acknowledged & used as a way
correlation and energy.
for authorization in our daily life, which is an important
 The objective of our systems is used to recognition
biometric attributes of human used to verify human
of signature by signers. In this phase, we have to
identification. Manual signature is fundamental
compare the texture features of test images with
procedure for individual, which is used for uncovering of
features of train images. If it’s matched then the
the document of signer with the assumption that the
given signature is identified else not.
signature varies slowly & virtually unfeasible to falsify
without detection. A signature is difficult to replicate and
broadly used to identify an individual delivering his day OVERVIEW OF SYSTEM
by day events such as document study, bank activities, Offline signature recognition system is an automated
electronic funds transfer and access control. A signature process of detection by extracting local features that
as a behavioral biometric encrypts the ballistic actions of classify each input signature. In this system, we initiate
an individual and allows higher intra-class and time by images are scanned using scanner, elaborated the
inconsistency, estimates the physical qualities which are input signature by preprocessing, the extraction of
fingerprint, iris or face. texture features from the preprocessed images and
analyze the signature with the signature stored in the
Depending on exhaustion, psychological and physical knowledge base using classification technique. If its
state, and lettering location (ergonomics), signatures match then input signature is recognized else not.
vary. The marker accelerations, which are comparative
to the muscle forces exerted by the signer, are reliable in
a usual signature.

Motivation
Offline signature recognition system is a significant
biometric technique, which is used to offers automated
process of recognition and authentication by extracting Fig. 1 System overview.

www.ijeid.com {IJEID © 2018} All Rights Reserved Page | 46


Artificial neural network based offline signature recognition system Rampur S.M.

PROPOSED METHODOLOGY and unwanted elements. To overcome the corruption


The proposed methodology consists of several essential caused due to noise, we use median filter for smoothing
steps which are supposed to be followed with précised and recovering images. Median filter is non-linear
rules. The block diagram of our proposed system gives operation which is used to reduce noise and preserve
the short idea about the procedure of methodology. edges. This conversion is shown in Figure 4.

Thinning
It is one of the morphological procedures, which is
constructed with procedures on pixel sets.
Morphological operations take two arguments: binary
image and structuring element. To conserve the aspect
ratio of signature image, thinned signature image goes
through normalization step. The thinned image consists
of 0’s and 1’s constituted by pixel sets, where the pixels
in the signature become less. This procedure is shown in
Figure 5.

(a) (b)
Fig. 2: Block diagram of proposed work.
Fig. 3: RGB form of image to gray-scale conversion. (a)
RGB form of image. (b) Gray-scale form of image.
The major steps of the signature recognition are
explained as follows:
Image Acquisition
The action of retrieving an image from the hardware based
source is known as image acquisition, so that it can pass
through the processes which occur after image acquisition.
The first step in the workflow sequence is image
acquisition, because without an image, processing is (a) (b)
impossible. Handwritten signatures are signed on the white
Fig. 4: Noise reduction of image. (a) image contains
paper by the individuals at different timing under different
noise. (b) noise reduction.
emotions such as stress and joy levels and these signatures
are scanned by using scanner of desired dpi resolution in
jpeg image format for the creation of the database.

Pre-Processing
It is a technique to augment raw images by removing
distortions and is the first part of the proposed system
prior to computational processing. We have used RGB- (a) (b)
to-grayscale conversion, gray scale to binary image
conversion, noise reduction, thinning and resizing for Fig. 5: Thinning of image. (a) binary form of image. (b)
pre-processing. thinned image.

RGB to Gray-Scale Conversion The Feature Extraction Phase


In this phase, if the original image is in the form of red, It is the most significant phase in any recognition system
green and blue which is used to convert to gray scale by because accuracy of the recognition completely depends
using following formula: on the features which are going to be extracted from the
Gray-scale = (0.299*R) + (0.5876*G) + (0.114*B) signatures. The main purpose of the feature extraction
method is used to get back the features accurately. Here,
It is a general performance in the image preprocessing we have to extract local texture features like entropy,
phase, since processing of a three channel signature correlation, contrast, homogeneity and energy.
(colored image) is slower than that of processing a single
channel signature (gray-scale image). This conversion is
shown in Figure 3.

Noise Reduction
A signature images are corrupted due to addition of
unwanted elements or noisy channels and also gets
ruined due to destructive effects caused from lighting Fig. 6: Horizontal division.

www.ijeid.com {IJEID © 2018} All Rights Reserved Page | 47


International Journal of Excellence Innovation and Development
||Volume 1, Issue 1, Nov. 2018||Page No. 046-049||

The network collects neurons from the input level and


the output of the network is displayed on an output layer.

RESULTS AND DISCUSSION


The experiment has been approved out so as to estimate
calculated system’s performance. The experiment has
been worked on the database which consists of classes of
95 persons and 10 signatures per individual class where
Fig. 7: Block Image divisions. 80 textual features were extracted from each block of the
signature image. The textual features are listed as:
Classification
 Energy
This is the step where we obtain approximate or
 Correlation
appropriate result. Classification will be executed on the
basis of features extracted in the feature space.  Entropy
Classification divides the feature into different classes  Homogeneity
based on decision rules. Signature stored in the  Contrast
Knowledge Base is subjected to classification technique.
We have used Artificial Neural network as classifier in Identification rate of this experiment is 85-90% which
our signature recognition system. In classification we satisfied our requirement. The performance was checked
compare the test images with trained data images in against number of person signature also. Initially we
order to classify test images. started with database having 20 persons. Gradually we
increased our database by 20 persons in each step. The
following table 1 shows that 20 persons with accuracy.

Table 1: Signatures with accuracy.


Persons signature Accuracy
20 99%
40 98.3%
60 97%
80 96%
95 95%

We have checked performance of 20 signatures of 20


persons; keep adding by 20 persons in each step. We
have trained 20 signatures per each person using feed
Fig. 8: Artificial neural network. forward back propagation algorithm. The performance
rate has checked against signature as shown in following
The computational model such as artificial neural charts:
networks is stimulated by the brain, which are capable of
recognizing a pattern and machine learning. The
artificial neural networks are generally organized as Performance Rate
interrelated "neurons" and by passing information
through the network the values from inputs can be 100
determined. In general, neural networks are organized
into three layers such as input, hidden and output layer. 80
An 'activation function' is held by interconnected ‘nodes’ 60
which are included in neural network. By means of the
'input layer', patterns are given to the network, which 40 Performance
correspond to one or more 'hidden layers’, by using a 20 Rate
system of subjective ‘connections’ the actual processing
is done and then the hidden layers connect to an 'output 0
layer' which gives output of the system shown in the
figure 8 above.

Many artificial neurons together compose an artificial


neural network and are correlated according to accurate Fig. 9: Performance rate.
network design. The goal of the neural network is to
convert the inputs into significant outputs. Here, we have The above graph shows as a number of individual
used an algorithm called feed forward back propagation increases, performance getting decreases.
in ANN for identification and confirmation of signatures
of individuals, where the simulated neurons are CONCLUSION
structured into layers and transmit their signals Offline signature recognition system is a significant
“forward” and the errors are transmitted “backward”. biometric technique, which is used to offers automated

www.ijeid.com {IJEID © 2018} All Rights Reserved Page | 48


Artificial neural network based offline signature recognition system Rampur S.M.

process of recognition and authentication by extracting Verification by the Tracking of Feature and Stroke
local features that classifies each input signature and has Positions,” Pattern Recognition, vol. 36, pp. 91–
many number of uses. The proposed system has local 101, 2003.
texture features and feed forward back propagation in an [6] Vamsi Krishna Madasu, “An Automatic Offline
artificial neural network classifier to identify signature. Signature Verification and Forgery Detection
By using training algorithm called feed forward back System”, University of Canberra, Pattern
propagation, we can train a neural network using Recognition Technologies and Applications:
extracted features from the signatures. The artificial Recent Advances, 63-89, 2004.
neural network has given expected results. Our proposed [7] M. Hanmandlu, M. H. M. Yusof, and V. K.
system exhibited 85-90%success rate by identifying the Madasu, "Off-Line Signature Verification and
signatures correctly that it was taught for purpose. Forgery Detection using Fuzzy Modelling”,
March-2005, vol. 38, pp. 341-356.
Few more efficient features can be included to improve [8] Johannes Coetzer, “Off-Line Signature
our system. The performance of the system can be Verification”, Journal, University of Stellenbosch,
increased by using additional features in dataset. We can April -2005, 45-90.
increase accuracy of identification by increasing [9] Juan J. Igarza , Inmaculada Hernáez, Iñaki
signature of trained images for each individual. Our Goirizelaia, Koldo Espinosa and Jon Escolar,
system gives 90-99% accuracy based on number of input “Off-Line Signature Recognition based on
signature. Dynamic Methods”, Dept. of Electronics and
Telecommunications, University of the Basque
REFERENCES Country Alameda Urquijo s/n, Bilbao, Spain
[1] J.P. Drouhard, R. Sabourin, M. Godbout, E48013, vol.5779,2005.
“Evaluation of a Training Method and of Various [10] Dakshina Ranjan Kisku, Phalguni Gupta and
Rejection Criteria for a Neural Network Classifier Jamuna Kanta Sing, “Fusion of Multiple Matchers
Used for Off-Line Signature Verification”, IEEE using SVM for Offline Signature Identification”,
Int’l Conf. Neural Networks, Orlando, Fla., June Communications in Computer and Information
26-July 2, pp. 294-4,299, 1994. Science, 2009, Volume 58, pp. 201-208.
[2] R. Baeza-Yates, G. Valiente, “An Image
Similarity Measure Based on Graph Matching”, Biography of Author
IEEE University of Chile, 2000.
[3] E. J. R. Justino, F. Bortolozzi, and R. Sabourin,
“Off-Line Signature Verification using HMM for
Random Simple and Skilled Forgeries”, Proc. 6th
Intl. Conf. On Document Analysis and
Recognition, 2001, pp. 450-453.
[4] Xiao, X. and Leedham, G (2002), “Signature Mr. Shivashankar M. Rampur
received the B.E and M.Tech degrees in Computer
Verification using a Modified Bayesian Network.
Science and Engineering from Basaveshwar Engineering
Pattern Recognition”, 2002, vol. 35, no. 5, pp.
College, Bagalkot, Karnataka in 2012 and 2016,
983-995.
respectively. Presently working as Assistant professor at
[5] B. Fang, C.H. Leung, Y. Y. Tang, K. W. Tse, P.
Brahmdevdada Mane Institute of Technology, Solapur.
C. K. Kwok and Y. K. Wong, “Off-Line Signature

www.ijeid.com {IJEID © 2018} All Rights Reserved Page | 49

Das könnte Ihnen auch gefallen