A Novel Detection Model for Indian Fake Currency Using Android

Prof. Shalini Puri1, Himanshu Sharma2

Assistant Professor, Dept. of Computer Engineering, Poornima College of Engineering, Jaipur, India1

Student, Dept. of Computer Engineering, Poornima College of Engineering, Jaipur, India2

ABSTRACT itself. No need of Internet which makes After

This paper tells about a new technique of Purchase cost to NIL.
detecting fake currency. As the Technology After the Demonetisation of old notes,
Enhancing, counterfeiting of money is becoming counterfeit of new currency notes of
common in the market with advance features denomination 500, 2000 become very popular
which makes difficult for common persons to find because features of new notes are not generally
difference between fake and real notes. known by the people.
Reserve Bank of India (RBI) is only responsible for One benefit of using an Android Phone is that
printing currency notes in India. Governor of RBI there is no requirement of using external camera
signs the note for its Genuity. With modern for scanning the note.
printers available at low cost, it becomes easy for
individuals to print or copy the note at their 2. LITERATURE SURVEY
home. Some differences still left which can be
The technique for detecting fake currency
identified by a professional person only. Our
methods based on bit rate reduction techniques.
proposed system helps to detect fake currency
This article introduces a new approach using
with low cost of buying, so common people can
cost-saving techniques to extract the most
buy this system.
important data from the images of fake bills by
applying edge detection algorithms. This
Keywords – Image Acquisition, Feature
methodology is proposed by Mohammad H.
Extraction, Image Pre-processing.
Alshayeji.
This technique describes an electronic-based
1. INTRODUCTION
blind-based Enhancement System. This paper is
As technology helping in printing of fake notes, it designed primarily to support them and make it
also helping in detecting these fake notes. easier for them to use the currency. Here we
Different Systems are available to identifying the offer Android based apps to recognize the
fake currency notes but all these are costly or currencies from different countries, as well as
used by Banks only. Common people who is the their denominations for the blind. Image
most common victim for this crime can not afford processing techniques such as download
these systems. Our proposed system uses the features and matching are used to determine the
technology which is nowadays commonly currency. The software runs on low-end
available among the people i.e. Android. smartphone. We provide voice mail as an input
Android Phones or Smartphones are now to launch apps and capture images. Then the
available with most of the people, our system is image is captured and compared to the test
an Android Application which gives User image. If the features of both pictures are
Interface and all processing is done in the App matched, the audio output is given to the user

25
about the currency name and the country in provided as a result. This technique Is proposed
which it belongs. Otherwise, an error message is by Nayana Susan Jose and Sherman Siby.

3. SECURITY FEATURES OF INDIAN BANKNOTE (Rs 2000 & Rs 500)

2. Intaglio Printing Letters ‘RBI’ and ‘2000’ on the left side of the

The representation of Mahatma Gandhi, the banknote. This element can be seen well under

Reserve Bank seal, certification and guarantee an amplifying glass.

proviso, Ashoka Pillar Emblem on the left, RBI 6. Security Thread
Governor’s mark are imprinted in intaglio i.e. in Windowed security thread with inscriptions
raised prints, which can be felt by touch, in Rs. ‘भारत’, ‘RBI’, and ₹2000 on banknotes, with a
10, Rs. 50, Rs. 100, Rs. 500, Rs. 2000 notes. colour shift. The thread colour changes from
3. Latent picture green to blue when the note is tilted.
On the front side of Rs. 10, Rs. 50, Rs. 100, Rs.
500, Rs. 2000 notes, a vertical band on the 9. Number Panel
correct side of the Mahatma Gandhi’s picture Number panel with numerals growing from small
contains an idle picture demonstrating the to big on the top left side and bottom right side.
separate denominational incentive in numeral.
13. Bleed Lines
The inactive picture is obvious just when the
Seven angular bleed lines on left and right side in
note is held on a level plane at eye level.
raised print (obverse). Year of printing of the note
5. Micro lettering on the left (reverse).
This component shows up between the vertical
band and Mahatma Gandhi picture. Micro

26
4. PROPOSED SYSTEM Segmentation involves the portioning of an
image or volume into distinct non-overlapping
regions in a meaningful way.
Image Acquisition Segmentation of ID Mark is done using Otsu
Method.
Image Pre-processing Segmentation of Bleed Lines is very important in
order to identify the denomination of the note.

Segmentation 4.4 Feature Extraction

After Segmentation, features are extracted. This
Feature Extraction step is very challenging in image processing. The
main aim is to analyse and identify the note
under conditions such as old notes, worn out
Comparison
notes.
For comparison phase also features are required
Decision Algorithm to be compared with the test image.

4.5 Comparison
Output The most important phase of the system is to
compare the image captured with the original
4.1 Image Acquisition image saved in the system.
Image Acquisition is the first step in
Image processing. In our system this step is taken 4.6 Decision Algorithm
by android smartphone. Camera of phone is used After comparing the images, it is very important
with its advance features to capture the image. to take decision whether the currency is real or
fake. Different Algorithms are present for taking
4.2 Image Pre-processing this decision.
In the first step of pre-processing, background is Algorithm used for taking decision is Keypoint +
removed from the image by Edge based Decision Trees:
Foreground Extraction. An approach that is probably much faster than
After removing background, Noise Reduction by the others is using semantic texton forests. This
Median Filtering is done because captured image involves extracting simple keypoints and using a
might get blurred. To preserve this Median Filter collection decision trees to classify the image.
is used. In a Median filter, a window slices along This is faster than simple SIFT keypoint matching,
the image, and the median intensity value of the because it avoids the costly matching process,
pixels within the window becomes the output and keypoints are much simpler than SIFT, so
intensity of the pixel being processed. keypoint extraction is much faster. However, it
Sharpening of image is now done and next step is preserves the SIFT method's invariance to
to convert the RGB image into Gray Scale Image, rotation, scale, and lighting, an important feature
so that processing can be done easily. that the histogram method lacked.

4.3 Segmentation 4.7 Output

27
Result of the system is display on the android [9] Binod Prasad Yadav, C. S. Patil, R. R. Karhe, P.
application itself. H. Patil, “Indian currency recognition and
verification system using image processing” 2014
5. CONCLUSION IJARCSSE
[10] D. Alekhya, G. Devisuryaprabha, G. Venkata
Here we used Android Application for user
Durga Rao “fake currency detection using image
interface and OpenCV Library for image
processing and other standard methods”
processing and MATLAB tools are also used in
[11] www.rbi.org.in
some steps of the system.
Main advantage of the system is the All at One
place methodology.

6. REFERENCES
[1] Ballado, A. H. Jr., Dela Cruz, J. C., Avendano,
G. O., Echano, N. M., Ella, J. E., Medina, M.E.M.,
Paquiz, B. K. C. “Philippine currency paper bill
counterfeit detection through image processing
using canny edge technology” 2015 IEEE
[2] Kamesh Santhanam, Siram Sekaran, Sriram
Vaikundam, Anbu mani Kumarasamy
“counterfeit currency detection technique using
image processing, polarization principle, and
holographic technique” 2013 IEEE
[3] M. Deborah, C. Soniya Prathap. “Detection of
fake currency using image processing” 2014
IJISET
[4] Ajinkya Babar, Swapnil Jawalekar, Kiran
Yadav, Dr. D. B. Salunke “counterfeit currency
detector” jun 2015 IJTRA
[5] Sagar S. Lawade, gayatri S. hedau, Apurva C.
Ramgirwar “fake currency detection using image
processing” IJRISE
[6] B. sai Prasanthi, D. Rajesh Shetty “Indian
paper currency authentication system using
image processing” IJRET 2015
[7] Snigdha Kamal, Simarpreet Sing Chawla, Nidhi
Goel, Balasubramanian Raman “Feature
extraction and identification of Indian currency
notes” IEEE
[8] Zahid Ahmed, Sabina Yasmin, Md Nahidul
Islam, Raihan Uddin Ahmed “Image Processing
based feature extraction of Bangladeshi
banknote” 2014 IEEE

28