You are on page 1of 2

Fake Currency Detection System – A Comprehensive

Review
Under The
Supervision of
Professor

Mr. Akhilesh Kumar Singh

Submitted By:
Hemant Chaudhary (21SCSE1180012)
Santrim Sharma (21SCSE1180259)
Satyam Nath Tiwari (21SCSE1180224)

Abstract literatures and identify the advantages and


disadvantage for each method in these literatures.
The currency has a great meaning in everyday The results were summarized in a comparison
life. Thus currency recognition has gained a great table that presented different ways of reviewing
interest for many researchers. The researchers the technology used in image processing to
have suggested diverse approaches to improve distinguish currency papers.Counterfeit notes are
currency recognition. Based on strong literature one of the biggest problem occurring in cash
survey, image processing can be considered as the transactions.
most widespread and effective technique of
currency recognition. This paper introduces some I. THE MODEL ARCHITECTURE
close related works of papercurrency recognition. This model we will build will use a grayscale
This paper has explained a variety of different image, segmentation, and feature extraction. We
currency recognition systems. The applications are going to use OpenCV in this project because
have used the power of computing to differentiate this library contains some of the pre-built
between different types of currencies with the functions to work with images. To check whether
appropriate layer. Choosing the proper feature the currency is fake or real, we will take a real note
would improve overall system performance. The and a fake note as input an image in the system.
main goal of this work is to compare previous The input images given to the system will be
papers and literatures through reviews these converted into grayscale images. We will use

XXX-X-XXXX-XXXX-X/XX/$XX.00 ©20XX IEEE


segmentation to segment out the image of Gandhi recognition using OpenCV python has notable
Ji and the thin strip on the note. demand in the current research community..

ACKNOWLEDGMENTS
The idea is that we will compare the real and fake Technology is growing very fast these days.
notes based on the image of Gandhi Ji and the thin Consequently, the banking sector is also getting
strip present on the image. We will compare the modern day by day. This brings a deep need for
image of Gandhi Ji and we will check whether the automatic fake currency detection in automatic
number of lines in the thin strip is equal to the teller machines and automatic goods seller
number of lines in the thin strip. machines. Many researchers have been encouraged
The comparison of Gandhi Ji’s image is simple to develop robust and efficient automatic currency
because we just have to find a correlation between detection machines. An automatic machine that can
two images using a correlation function. If the detect banknotes is now widely used in dispensers
result of the correlation function is greater than 0.5 of modern products like candies, soft drinks bottles
then Gandhi Ji is legitimate and we will check for buses, es or railway tickets. The technology of
a thin strip. Otherwise, the note is fake. Now the currency recognition basically aims for identifying
main problem is to find the number of lines in thin and extracting visible and invisible features of
strips. For that, we will convert the image into currency notes. Until now, many techniques have
HSV(hue saturation value) image because it will been proposed to identify the currency note. But the
be easy that way. Then we will segment the image best way is to use the visible features of the note.
For example, color and size. But this way is not
to find the thin strip. After that, we will apply a
helpful if the note is dirty or torn. If a note is dirty,
bwlabel function to find the number of connected its color characteristic is changed widely. So it is
components i.e. the number of lines in the thin important that how we extract the features of the
strip. image of the currency note and apply the proper
In the end, we will just write small code to final algorithm to improve the accuracy to recognize the
check whether the currency is fake or real. note. With this Machine Learning Project, we will
be building a fake currency detection system for
NOTE: We have just used two features i.e. Gandhi detecting fake currency and real currency. In
Ji image and thin strip. But there are a lot of other today’s time, the production and circulation of
things on a real note like a unique number, bleed inauthentic currency notes have been increasingly
lines(using which a blind man can find what is the sophisticated. That’s why there needs to be a system
value of the currency), etc. For improving the to check whether a currency is fake or real.
model, you can try implementing these as
“OpenCV”. This library is mainly intended for REFERENCES
image processing, computer vision/graphics, and 1.https://ieeexplore.ieee.org/document/8777621
machine learning fields. For this purpose, it
comprises nearly about 2000+ optimized 2.https://www.pantechelearning.com/product/currenc
algorithms. Further, it supports various y-detectionusing-machine-learning-opencv-and-
programming languages such as Java, Python, python/
C++, etc. As a matter of fact, it can detect a human
face, handwriting, and other objects in 3.https://www.researchgate.net/publication/32273873
images/videos.By the by, it can process images 1_Paper_Currency_Detection_based_Image_Process
and videos to identify objects, faces, or even the ing_Techniques_A_revie
handwriting of a human. For advanced processing, w_paper/link/5a6c4971a6fdcc317b15f4a9/download
several packages like NumPy, matplot, etc. van be
4.https://github.com/aprameya2001/Fake-Currency-
incorporated. Overall, python enables to process of
OpenCV array structure for comprehensive Detection-System
analysis. Further, it also allows working various
5.https://www.jagranjosh.com/general-
mathematical operations using vector points on knowledge/security-features-of-indian-currency-
essential image patterns/features.Overall, currency notes-1531736583-1

You might also like