Counterfeit Currency Detection

@ IJTSRD | Available Online @ www.ijtsrd.com
ISSN No: 2456
International
Research
Counterfeit Currency Detection
Madhuri R.
Department of Electronics and Telecommunication
MAEER’S MIT College of Engineering, Pune,
ABSTRACT
Nowadays problem of fake currency increases because
of increasing in technology like scanning, color
printing so result in counterfeit currency. In India
increase in fake paper currency notes of 100, 500, 2000
rupees etc. So detection of fake currency is n
The determination of fake currency with the help of
image processing. Firstly Image acquisition is done
then pre-processing stage applied to that image for
suppress unwanted feature and added some feature
which are necessary for further process. Conversion of
RGB picture into HSV scale. Then image segmentation
applied to that image in this image divided into number
of objects. Then morphological operation is perform on
that picture. Further feature extraction/area of
calculation stage applied to that picture and finally that
picture compared with the original image.
Keywords: Fake currency, image processing, feature
invarient, counterfeit, HSV
I. INTRODUCTION
Many times RBI faces problem of counterfeit currency
that major consequences on Indian economy so also
increases additional problem. So adding such a
technology or machine which will make human efforts
more simpler and efficient. For the detection of
counterfeit currency bank employee always keep note
on that device and tries to find all the feature of
@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 4 | May-Jun 2018
ISSN No: 2456 - 6470 | www.ijtsrd.com | Volume
International Journal of Trend in Scientific
Research and Development (IJTSRD)
International Open Access Journal
Counterfeit Currency Detection
Madhuri R. Raut, Prof. Dr. Krishna K. Warhade
Department of Electronics and Telecommunication
MAEER’S MIT College of Engineering, Pune, Maharashtra, India
Nowadays problem of fake currency increases because
of increasing in technology like scanning, color
printing so result in counterfeit currency. In India
increase in fake paper currency notes of 100, 500, 2000
rupees etc. So detection of fake currency is necessary.
The determination of fake currency with the help of
image processing. Firstly Image acquisition is done
processing stage applied to that image for
suppress unwanted feature and added some feature
Conversion of
RGB picture into HSV scale. Then image segmentation
applied to that image in this image divided into number
of objects. Then morphological operation is perform on
that picture. Further feature extraction/area of
hat picture and finally that
Fake currency, image processing, feature
Many times RBI faces problem of counterfeit currency
dian economy so also
increases additional problem. So adding such a
technology or machine which will make human efforts
more simpler and efficient. For the detection of
counterfeit currency bank employee always keep note
l the feature of
currency and finally identification of fake currency so
workload of bank employee increase. So result could more
accurate using this system. This system also very useful in
shopping malls and investment firms. Automatic
counterfeit currency discovery is vital
application as automatic seller goods machine and
automatic teller goods machine. This system is useful for
detection of fake Indian currency. Eight stages associated
with discovery of fake note, for example picture obtaining,
data scale transformation, edge identification, highlight
extraction, picture division examination of info and yield.
This system useful in bank as well small shops that face
more problems of counterfeit currency.
Commonly used methods to detect fake no
For 2000 note:
Front side features:
 can see numeral 2000 when note held against the light
 At the point when note will be worked then the
dormant picture of 2000 will happen.
 2000 and ‘RBI’ having the color shift security thread
 Number board with numbers develop.ping from little
to huge on the upper left to base right sides
 Ashoka pillar emblem
Jun 2018 Page: 1251
www.ijtsrd.com | Volume - 2 | Issue – 4
Scientific
(IJTSRD)
International Open Access Journal
India
currency and finally identification of fake currency so
workload of bank employee increase. So result could more
accurate using this system. This system also very useful in
shopping malls and investment firms. Automatic
y discovery is vital in numerous
as automatic seller goods machine and
teller goods machine. This system is useful for
detection of fake Indian currency. Eight stages associated
with discovery of fake note, for example picture obtaining,
scale transformation, edge identification, highlight
extraction, picture division examination of info and yield.
This system useful in bank as well small shops that face
more problems of counterfeit currency.
Commonly used methods to detect fake notes:
can see numeral 2000 when note held against the light
At the point when note will be worked then the
dormant picture of 2000 will happen.
2000 and ‘RBI’ having the color shift security thread
rd with numbers develop.ping from little
to huge on the upper left to base right sides

International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 4 | May-Jun 2018 Page: 1252
Fig.1 Security features of 500 & 2000 note
Backside feature:
 We can see Swachh Bharat logo on back side of
the note
For 500 note:
Front side features:
 We can see numeral 500 when note held against
the light
 When note will be tilled then the latent picture of
500 will be happen
 The orientation of Mahatma Gandhi portrait
changed
 Number panel with number changes from little to
huge on the upper left to right side.
Backside feature:
 We can see Swachh Bharat logo on back side of
500 note
 Red Fort and Indian flag image
 Rs 500 note having circle in Devnagari on the
right side
II. LITERATURE SURVEY
As we know the Printing house being able to make
fake paper currency but it is workable for any
individual to print fake cash simply utilized for
computer and laser printer at home. Fake currency
detection methods are explain is as below
Trupti Pathrabe and swapnili karmore [3] introduced a
new technique for the classification of Japanese and
U.S paper currency that will improve the
identification ability and transaction speed. Trupti
Pathrabe and Swapnili analyze two sorts of
informational indexes which include the time series
data and Fourier power spectra. The neural framework
is used as input for two cases. For recognition ability
they used new evolution method In this system
Embedded system is used for the detection of fake
note. This system emphasis on Indian currency
features. So no chance of making counterfeit note by
people.
The recognition system divided into three parts firstly
include the MATLAB is used for reduce image size
and extract feature of scanned picture that will used
for advance process.HSV shading space completed
the work of color feature extraction. Second stage
acknowledgement is neural network classifier and
lastly fake currency detection will be visible on AVR
microcontroller ATMmega32.The AVR
microcontroller that determines the validity of given
note by glowing the LED light for counterfeit Indian
currency. Pictorial information changes by human
interpretation with image processing. With the help of
the image processing tools collection functions that
increase the capacity of MATLAB. That mainly
emphasis on HSV color space.
RGB shading space and HSV shading space are not
same in light of the fact that the RGB shading space
isolates the power of that photo from the shading data
Future work include determination of foreign
currency with the same system.
Algorithm used by Komal Vora[1] based on
frequency domain feature extraction. This method
efficiently used the local spatial feature in a currency
picture to identify it. Human cannot recognize
currency of different countries easily .In this method
textural feature are extracted from the Discrete
wavelet transform. Non-textural feature are utilize for
checking authenticity. They are such a serial number,
shading and the textural feature used to group them.
Pattern matching gives the required output.Because of
that denomination validation will be finished. so that
authentication process check the feature of currency
like security thread, RBI microprint and then identify
currency is fake or real. Future work will be
conversion of currency denomination.
Pathrabe and Bawane [5] use a low computational
complexity algorithm that uses this author to meet
high-speed practical requirements application. It
needs to be proposed system has to distinguish
between fake and genuine currency. And this is
cannot identify counterfeit and genuine currency.
Infrared or ultraviolet spectra is used by this technique
for determination of counterfeit and original currency.
In last decade problem of counterfeit currency
increases this is only because of scanning, color
printing. Few years ago only printing house able to

make fake paper currency but nowadays any person
can make currency simply by using computer and
laser printer. Proposed system given by the Rajesh
Shetty and Sai Prasanthi. This system having
advantage of simplicity and high speed. Here image is
taken by the camera by applying backlighting to that
paper currency so that hidden feature see on the
picture. Picture then further processed by applying
image processing techniques. In this process
characteristic extraction plays very important role
because it have to extract visible and invisible feature
of paper currency. Good characteristic extraction
should maintain and enhance the some feature of
input picture which help to distinct pattern classes
separate from each other. Future scope the system will
extract feature even the testing image having the
different size [6].
Extract feature from the banknote denomination
technique used by Mirza and Nanda [4]. The region
will be extracted from the image will be used for
recognition of pattern and matching technique of
neural network. In which firstly image acquired by the
simple scanner on fix dpi with the particular size in
which pixel level has been set for obtaining an image.
For denomination value of note few filters are
applied. They used different pixel level in different
denomination note. For finding the denomination
value the pattern recognition and neural matching
technique used. The three characteristics of Indian
paper money, such as security thread, identification
mark and watermark, are selected for the counterfeit
currency.
Then characteristic of the image of paper currency
that will compare with genuine currency. For
characteristics extraction, Sobel operator is use. This
system has the main advantage is simplicity and high
speed. In future they will work on hardware part
which will used for picture acquisition for minimizing
the counterfeit currency.
Proposed system
Fig.2 Block diagram of counterfeit currency
detection
1) Input Image: Here image will be taken by the
camera. The acquired picture consists of all the
features.
2) Image Pre-processing: The point of picture
processing is remove undesirable/undesired
distortion and enhance some photo feature that are
valuable for additionally process .In that it Include
Picture adjusting and picture smoothing.
3) Color space conversion: The image obtained from
the camera is in a RGB scale so need to convert
that RGB scale into HSV image because that
scale image contain all the intensity information
that is necessary for further process.
4) Image segmentation /thresholding: Picture
thresholding is a basic, yet successful, method for
dividing a picture into a forefront and
background.This pictureexamination procedure is
a kind of picture segmentation that isolates objects
by changing over HSV images into binary
pictures. Image thresholding is most impressive in
picture with high levels of contrast.
5) Morphological operation: Morphology is a wide
arrangement of picture processing operations that
procedure pictures based on shapes.
Morphological tasks apply a structuring
component to data picture, making an output
photo of the similar size. In a morphological
movement, the value of every pixel in the output
picture is relies upon a correlation of the
corresponding pixel in the information picture
with its neighbors. By picking the size and state of
the neighborhood, you can develop a
morphological task that is touchy to specific state
in the information picture. The most basic
morphological task are enlargement and
disintegration. Enlargement adds pixels to the

limits of items in a picture, while disintegration
evacuates pixels on question limits. The quantity
of pixels included or expelled from the things in a
photo relies upon the size and conditon of the
organizing component used to process the photo.
In the morphological widening and disintegration
activities, the condition of any given pixel in the
yield picture is dictated by applying an administer
to the comparing pixel and its neighbors in the
information picture. The run used to process the
pixels characterizes the task as an enlargement or
a disintegration.
6) Feature extraction: Feature extraction a kind
dimensionality decrease the efficiently represents
interesting parts of picture as a reduce feature
vector. This approach is valuable when picture
sizes are enormous and a diminished element
portrayal is important to rapidly total work, for
example, picture coordinating and recovery.
Feature recognition, feature extraction, and
matching are many times combined to take care of
normal PC vision issue for example, object
detection, recognition, content-based image
retrieval, face detection and recognition, and
texture classification. The size of picture will be
easily reduced by feature extraction. Input given
to the algorithm is too large for further process it
is having much data but not more information.
Then input data will be converted into reduced
representation set of features
7) Comparison: In this stage, the extracted feature of
input image and extracted feature of original
image is compared.
III. RESULTS
The results of the proposed method are described
below:
(a)
(b)
(c)
(d)
Fig. 1. The results of proposed system on real currency
(a) Input Image (b)HSV conversion (c) Segmentation
results (d) Morphological operation
(a)

(b)
(c)
(d)
Fig.4.The results of proposed system on Fake
currency (a) Input Image (b)HSV conversion (c)
Segmentation results (d) Morphological operation
From the results of the proposed system it is observed
that, the features are appeared in the real currency
while the selected features are vanish in fake
currency. In this approach finally the number of white
pixels in binary image is calculated and make a
decision
V.CONCLUSION
Detecting of fake currency is necessary for comman
man today. So, In this project, detection of fake
currency will be done by image processing principle.
This is low cost system. The system will work for
note denomination of 500, 2000 100,200 currency etc.
The system also provide accurate and valid results.
The process of detection of fake note will be quick
and easy. In this input will be taken by CCD camera
output will be displayed on PC
REFERENCES
1) Komal Vora, Ami Shah, Jay Mehta, “A Review
Paper on Currency Recognition System”,
International Journal of Computer Applications,
vol.115,no.20,pp.125-130 ,2015.
2) Prof. Renuka Nagpure, Shreya Shetty, Trupti
Ghotkar, “Currency Recognition and Fake Note
Detection”, IIRC, vol-4,Issue 3, pp.3659-
3666,2016
3) G. Trupti Pathrabe, Mrs.Swapnili Karmore, “A
Novel Approach of Embedded System for Indian
Paper Currency Recognition”, International
Journal of Computer Trends and Technology,
pp.152-156,2010
4) Rubeena Mirza, Vinti Nanda, “Characteristic
Extraction Parameters for Genuine Paper
Currency Verification Based on Image
Processing”, IFRSA International Journal of
Computing, vol. 3,Issue 2,pp.41-46 2012.
5) Pathrabe T, Bawane N.G, “Feature Extraction
Parameters for Genuine Paper Currency
Recognition Verification”, International Journal of
Advanced Engineering Sciences and
Technologies, vol 2, pp.85-89, 2011.
6) B.Sai Prasanthi, D. Rajesh Setty, “Indian Paper
Currency Authentication System using Image
processing”, International Journal of Scientific
Research Engineering Technology (IJSRET),
vol.4, Issue 9,pp.973-981,2015
7) Vibhuti Vashishtha and Md. Sadim “A Paper
Currency Recognition System Using Image
Processing To Improve the Reliability of PCA
Method”. International Journal of Engineering
Science Research Technology. vol-1,no.
1,pp.2277-2280, 2015
8) Sonali R. Darade, Prof. G.R.Gidveer2, “
Automatic Recognition of Fake Indian Currency
Note”, International Research Journal of
Engineering and Technology (IRJET), vol: 03
Issue: 12,pp. 415-418,Dec-2016

Counterfeit Currency Detection

More Related Content

What's hot (20)

Similar to Counterfeit Currency Detection (20)

More from ijtsrd (20)

Recently uploaded (20)

Counterfeit Currency Detection