Fake Currency Detection Using

Image Processing Techniques

A Mini Project report submitted
in the partial fulfillment of the requirements for the award of the degree of

Bachelor of Technology
in
Computer Science & Engineering
(Internet of Things)
By
Roll No-20071A6928 Name-K.Deeksha

Under the Guidance of

Dr. Sagar Yeruva
Associate Professor
Submitted to

DEPARTMENT OF
CSE- ( AIML & IoT)
Vallurupalli Nageswara Rao Vignana Jyothi Institute of Engineering and Technology,
Hyderabad, Telangana
September, 2023
 VALLURUPALLI NAGESWARA RAO VIGNANA JYOTHI
INSTITUTE OF ENGINEERING AND TECHNOLOGY
An Autonomous, ISO 9001:2015 & QS I-Gauge Diamond Rated Institute, Accredited by NAAC with ‘A++’ Grade
NBA Accreditation for B.Tech. CE, EEE, ME, ECE, CSE, EIE, IT Programmes
Approved by AICTE, New Delhi, Affiliated to JNTUH, NIRF 135th Rank in Engineering Category
Recognized as “College with Potential for Excellence” by UGC
VignanaJyothi Nagar, Pragathi Nagar, Nizampet (S.O), Hyderabad – 500 090, TS, India.
Telephone No: 040-2304 2758/59/60, Fax: 040-23042761

CERTIFICATE

This is to certify that K.Deeksha (20071A6928) have successfully completed their Mini project work
at CSE(AIML & IoT) Department of VNRVJIET, Hyderabad entitled “Fake Currency Detetction
Using Image Processing Techniques” in partial fulfilment of the requirements for the award of B.
Tech degree during the academic year 2022-2023.
This work is carried out under my supervision and has not been submitted to any other
University/Institute for award of any degree/diploma.

GUIDE Head of the Department

Dr.Sagar Yeruva Dr. N. Sandhya
Assosiate Professor Professor & Head
Dept. of CSE (AIML & IoT)
Dept. of CSE (AIML &IoT)

DECLARATION

1
 This is to certify that our project titled “Fake Currency Detection Using Image Processing
Techniques” submitted to Vallurupalli Nageswara Rao Vignana Jyothi Institute of Engineering and
Technology in complete fulfilment of the requirement for the award of Bachelor of Technology in
CSE (AIML & IoT) is a bonafide report to the work carried out by us under the guidance and
supervision of Mr. Sagar Yeruva, Assosiate Professor, Department of CSE(AIML & IoT), Vallurupalli
Nageswara Rao Vignana Jyothi Institute of Engineering and Technology. To the best of our
knowledge, this has not been submitted in any form to another University/Institute for an award of
any degree/diploma.

K.Deeksha
20071A6928
Dept. of CSE (IoT)

2
 ACKNOWLEDGEMENTS

I owe an inestimable debt of gratitude to my guide Dr.Sagar Yeruva , Assosiate Professor

Department of Computer Science and Engineering (AIML & IoT) Vallurupalli Nageswara rao

Vignana Jyothi Institute of Engineering and Technology, Hyderabad for his/her help and guidance in

successful completion of this dissertation. They were always there to answer my questions and set

time to meet me whenever needed. I will be forever grateful to them.

I would like to express our gratitude towards the Principal of our institute Dr.Challa

Dhanunjaya Naidu and Head of the Department Computer Science and Engineering (AIML &

IoT),Dr.N.Sandhya for their kind cooperation and encouragement which helped us to complete the

project in the stipulated time.

3
 ABSTRACT

Fake currency is the money produced without the approval of the government, creation of it is
considered as a great Offence. The elevation of colour printing technology has increased the rate of
fake currency note printing on a very large scale. Years before, the printing could be done in a print
house, but now anyone can print a currency note with maximum accuracy using a simple laser
printer. This results in the issue of fake notes instead of the genuine ones has been increased very
largely. It is the biggest problem faced by many countries including India. Though Banks and other
large organizations have installed Automatic machines to detect fake currency notes, it is really
difficult for an average person to distinguish between the two[1].

This has led to the increase of corruption in our country hindering the country's growth. Some of the
methods to detect fake currency are watermarking, optically variable ink, security thread, latent
image, techniques like counterfeit detection pens. We hereby propose an application system for
detecting fake currency where image processing is used to detect fake notes. We are going to detect
the variation in barcode among the real and fake one and also, we will find out dissimilarities
between the image under consideration and the prototype. Image processing technqiues will be
used to detect fake currency. The proposed web application for fake currency detection will be
simple, accurate and easy to use[6].

4
 TABLE OF CONTENTS

1. Table of Contents 1
2. Table of Figures 4
3. List of Tables 5

1 INTRODUCTION 5

1.1 EXISTING SYSTEM: 7

1.2 PROPOSED SYSTEM: 7

2 LITERATURE SURVEY 9

2.1 RELATED WORK 9

2.2 SYSTEM STUDY 12

3 DESIGN 15

3.1 SYSTEM REQUIREMENTS 15

3.1.1 Software Requirements 15

3.1.2 Hardware Requirements 15

3.2 UML 16

3.2.1 Use case 16

3.2.2 Class Diagram 17

3.2.3 Activity Diagram 17

3.2.4 Statechart 18

4 IMPLEMENTATION 15

4.1 MODULES 15

4.1.1 Dataset 20

4.1.2 Models 21

4.1.3 Performance Metrics 21

4.2 OVERVIEW TECHNOLOGY 22

4.2.1 Image Processing Techniques 22

4.2.2 TKinter 23

4.2.3 Python 23
5
5 TESTING 24

5.1 TEST CASES 24

6 RESULTS 25

7 CONCLUSION 28

8 FUTURE SCOPE 29

9 BIBILOGRAPHY 30

6
 LIST OF FIGURES

Figure 1-1:It shows manual way of detecting fake or real note 5

Figure 1-2 : The number of fake notes in detected in circulation over the last three years 6
Figure 1-3: Block Diagram 7
Figure 3-1: Use Case Diagram 16
Figure 3-2 : Class Diagram 17
Figure 3-3 : Activity Diagram 18
Figure 3-4: State Chart Diagram 19
Figure 6-1: Website 25
Figure 6-2: Inserting 500 rupee note 26
Figure 6-3: All test cases are passed 26
Figure 6-4: 2000 rupee note sent for processing 27
Figure 6-5: All features are not passed 27

7
 LIST OF TABLES

Table 5.1-1 Test Cases 24

8
 1 INTRODUCTION
[1][6]Counterfeit currency production is a pressing issue faced by countries globally,
undermining the value and stability of real money while potentially leading to inflation
through unauthorized money supply increase. Manual authentication of currency notes is
impractical for handling large volumes, being time-consuming and error-prone. Addressing
this concern, a crucial solution lies in the implementation of an automatic testing system for
currency notes, utilizing image processing techniques and algorithms for efficient and
accurate detection of fake currency.

In this proposed project, a dedicated fake currency note detection system is designed
to validate Indian currency notes of 100,500 and 2000 rupees denominations. The system
incorporates three primary algorithms, each focusing on verifying specific features within a
currency note. The first algorithm encompasses image acquisition, pre-processing, greyscale
conversion, feature extraction, image segmentation, and comparisons, leveraging advanced
image processing methods like ORB and SSIM to validate the authenticity of various features.
The second algorithm concentrates on authenticating the bleed lines of the currency notes,
while the third algorithm assesses the number panel. The resulting automated system offers a
seamless way to authenticate currency notes swiftly and accurately, presenting a viable
alternative to current manual methods and ensuring wider usability for fake currency
detection[4][5].

Fig1:It shows manual way of detecting fake or real note .

As shown in the figure given above traditional methods are time-consuming and not giving
accurate results compared to automated methods.

9
By integrating image processing algorithms and techniques, this automated fake currency note
detection system offers a streamlined approach to address the rising concerns of counterfeit currency.
The system's efficiency in authenticating Indian currency notes of specified denominations – 100 ,
500 and 2000 rupees - demonstrates its potential to replace time-consuming and inaccurate manual
methods. Ultimately, this automated solution significantly contributes to safeguarding the integrity of
the monetary system by swiftly identifying and mitigating counterfeit currency, thereby preserving
the value and stability of real money.preserving the value and stability of real money[1][6].

Fig 1.2: The number of fake notes in detected in circulation over the last three years

Fake currency has been a thorn in India’s economic system even after the Narendra
Modi administration’s demonetisation exercise back in 2016. Four years down the line, and
after adding ‘security features’, counterfeit currency still seems to be lurking in the shadows.
Overall, there has been a 29.7% decrease in the circulation of fake currency over the
last one year. However, at the same time, the number of fake ₹500 notes in India jumped by
31%, according to the Reserve Bank of India’s (RBI) latest annual report.
While other denominations have been seeing a drop in the number of fake notes
circulated, the figures released by the RBI do not include the counterfeit notes seized by the
police and other law enforcement agencies[6].

10
 1.1 EXISTING SYSTEM
In the current situation fake note detection primarily relies on traditional models that
have evolved over many years. These systems are characterized by several key
components:
In-Person Inspection: The predominant method of fake note detection is physical
inspection of various security features that are difficult for counterfeiters to replicate.
Lack of Automation: Traditional methods do not leverage automation or technology,
making them less efficient and unable to handle large volumes of currency quickly[1].

Disadvantages of existing systems:

Dependency on MATLAB:Some of the systems rely on MATLAB for

implementation, which may limit accessibility and ease of use for individuals and
businesses without MATLAB expertise. MATLAB can also be a paid software, adding
to the cost.
Limited Denominations: Most of the systems are designed to detect specific
denominations of currency notes. This limitation makes them less versatile and
applicable to a broader range of currency denominations.
Lack of Real-Time Processing: It's mentioned that the notes are captured by a
camera and then analyzed using software. This process may not be real-time, making it
less suitable for situations where quick authentication of currency notes is required.
While traditional methods have been valuable for many years, their limitations have
prompted the development of more advanced and automated counterfeit detection
systems that leverage technology, image processing, and machine learning to provide
more accurate and efficient results[8][9].
1.2 PROPOSED SYSTEM:

Figure 1-3 Block Diagram

11
 This Figure shows the steps in the proposed workflow which involves the pre-
processing of image, grey scale conversion, and feature extraction. with specified image
processing algorithms and evaluation of results and prediction of note genuity[2].
The system consists of three main algorithms and checks the authenticity of various features in a
currency note as follows:
Algorithm 1:
In this algorithm, we're using a method called ORB to detect important security features in a currency
note image. These features are already known to us and are stored in our dataset. What ORB does is
help us pinpoint where these features are likely to be on the test image. We define specific search
areas in the test image where each feature is expected to appear, making the search more accurate.
Once we find these areas, we cut them out from the image and make them black and white
while also smoothing them slightly for better analysis. After this, we compare these cut-out parts
with the original feature templates using a scoring system called SSIM (Structural Similarity Index).
SSIM helps us quantify how similar the two images are. A higher SSIM score indicates a closer
match.
We calculate the SSIM score for each feature and find the average score. If the average
SSIM score is high, it means the features on the test note closely resemble the real ones. This
comparison of features allows us to determine whether the currency note is genuine or counterfeit
based on how well its features match with those in our dataset. This method helps automate the
process of currency authentication using image processing techniques[1][6][9].
Algorithm 2:
This algorithm is designed to identify and count the bleed lines on the left and right sides of currency
notes, specifically targeting 100,500 and 2000 rupee denominations.
Feature Extraction: Initially, we extract the areas near the note's edges where the bleed lines are
located, focusing on the left and right sides.
Image Thresholding: We then simplify the image by making it black and white, ensuring that only the
black bleed lines are visible against a white background.
Counting Bleed Lines: The algorithm counts the number of bleed lines present. It does this by
scanning each column of the thresholded image and checking how many black regions appear in each
column. If a column has too many black regions (10 or more), it's considered erroneous and
disregarded. The final count is the average number of non-erroneous columns, indicating the number
of bleed lines. Typically, this count should be around 5 for 500-rupee notes and 7 for 2000-rupee
notes[1][11].

12
Algorithm 3:
This algorithm is designed to count the number of characters in the serial number panel found on the
bottom right of currency notes, where there should be exactly 9 characters (excluding spaces).
Image Thresholding (with Variations): It starts by simplifying the image to make black characters
stand out on a white background using different threshold values, ranging from 90 to 150.
Contour Detection: The algorithm identifies the outlines (contours) of the characters in the
thresholded image.
Bounding Rectangles: It then draws rectangles around these contours to measure each character's
position and size.
Eliminating Unwanted Rectangles: Rectangles that are either too big, too small, or within larger
rectangles are removed to get rid of unnecessary noise.
Counting Characters: The remaining rectangles correspond to individual characters. The algorithm
counts these rectangles to determine the number of characters detected. It repeats this process with
different threshold values until it finds nine characters in three consecutive iterations or reaches
maximum threshold value of 150[1][11].

2 LITERATURE SURVEY

2.1 RELATED WORK

We referenced eight distinct research papers encompassing the detection of fake note
using image processing techiniques.
 FAKE CURRENCY RECOGNITION SYSTEM FOR INDIAN NOTES USING
IMAGE PROCESSING TECHNIQUES
Author: T Naveen Kumar, T Subhash, SK Saajid Rehman, N Hari Babu, P Sai, Dr.
D. Regan .
Publication Year: 2019
Published In: Journal of Emerging Technologies and Innovative Research
Methods Used: It uses DTCWT (Dual Tree Complex Wavelet Transform) are applied
on the pre-processed images.
Results: The basic image properties are used to evaluate the classification results in
the system. Matlab 2013b used for the simulation of the proposed system on windows
10 operating system with 4 GB RAM and 2.95 GHZ[2].

 Fake Indian Currency Recognition System by using MATLAB

13
 Author:Shivaji University, Kolhapur
Publication Year: 2019
Published In: International Journal for Research in Applied Science & Engineering
Technology (IJRASET)
Methods Used: MATLAB, entropy, hyper spectral imaging,
Results: Helps to classify the given note is 100,500 or 2000[3].
 Detection of Fake Currency using Machine Learning
Author : Pushpa R N1, Aditya Aithal H K2, Ganesh Prasad M3, Manoj R Divakar4,
Hithesha H G5
Publication Year: 2023
Published In: International Journal for Research in Applied Science & Engineering
Technology (IJRASET)
Methods Used: Machine Learning, Image processing, Convolutional Neural Network.
Results: The proposed approach can be highly beneficial for banks, businesses, and
individuals in detecting fake currency[4].
 Fake Currency Detection App
Author: Ravindra R. Ghugare 1, Prathmesh Shendge2, Dinesh Takale3,Rohit J.
Yadav4
Publication Year: 2021
Published In: International Research Journal of Engineering and Technology (IRJET)
Methods Used: It uses a sample dataset called indian Cuurency notes and the
algorithm used here for performing the action to detect the fake note is CNN.
Results: The application is designed for 25GB and above storage phones and the
designed app was little bit of clumsy .However it is able to detect the conterfiet
notes[5].

 APPROACH FOR HYBRID FAKE CURRENCY DETECTION SYSTEM

Author: Rikhitha.N1 , Hemanth Kumar B.N2
Publication Year: 2022
Published In: International Advanced Research Journal in Science, Engineering and
Technology
Methods Used: Image processing techniques used here done in wideranges for
obtaining the accurate results. The system is better trained for obtaining accurate
results.

14
 Results: The web application designed was able to generate accurate results but a
person with prior knowledge cant access the application generated[6].
 Fake currency detection using image processing
Author: Tushar Agasti
Publication Year: 2020
Published In: IOP Conference Series
Methods Used: MATLAB software is used to extract the features of the note. The
proposed system has got advantages like simplicity and high performance speed.
Results: The fake currency detection using image processing was implemented on
MATLAB. Features of currency note like serial number, security thread, Identification
mark, Mahatma Gandhi portrait were extracted. The process starts from image
acquisition to calculation of intensity of each extracted feature. The system is capable
of extracting features even if the note has scribbles on it. The algorithm processed here
works suitably for the newly introduced 500 and 2000 denomination[7].
 Detection of counterfeit indian currancy notes
Author: S Naresh Kumar, Gaurav Singal, Shwetha Sirikonda and R. Nethravathi .
Publication Year: 2018
Published In: IOP Conference Series
Methods Used: Algorithm used in this system is CNN.
Results: The paper presents a low-cost model for detecting features and verifying the
legitimacy of Indian currency notes, achieving an accuracy of 96.6% and successfully
handling scribbled notes. It highlights the importance of selecting features and tools
effectively to address this critical task[8].
 Research on Fake Indian Currency Note Detection using Image Processing
Author: Miss.I. Santhiya Irulappasamy
Publication Year: 2021
Published In: ijsdr
Methods Used: Using algorithms Structure Similarity Index Metric (SSIM), Adaptive
histogram equalization, Fast Discrete Wavelet Transform (FDWT) and Gray Level Co-
occurrence Matrix (GLCM). SSIM is a method used to find the currency is real/fake.
Results: In this project, the Structural Similarity Index (SSI) method is used to
distinguish genuine from counterfeit Indian rupee notes. Features are extracted using
Fast Discrete Wavelet Transform (FDWT) and Gray-Level Co-occurrence Matrix
(GLCM), and an Artificial Neural Network is employed for classification based on
features[9].
15
 2.2SYSTEM STUDY

 FAKE CURRENCY RECOGNITION SYSTEM FOR INDIAN NOTES USING

IMAGE PROCESSING TECHNIQUES
The research presented in the above document focuses on the development of a fake currency
recognition system for Indian banknotes using digital image processing techniques.The study aims to
address the significant issue of counterfeit currency circulation in India by leveraging image
processing algorithms to detect fake notes based on their physical characteristics, such as security
features. The proposed system involves image preprocessing, segmentation, and classification using
techniques like Dual Tree Complex Wavelet Transform (DTCWT) and k-means clustering, ultimately
providing a cost-effective and efficient means of identifying counterfeit currency[2].
 Fake Indian Currency Recognition System by using MATLAB
The system study described in the provided research project focuses on developing an
automated system for the recognition of counterfeit Indian currency using MATLAB. The study
involves two primary methodologies: hyperspectral imaging and image processing. Hyperspectral
imaging utilizes various light sources with different wavelengths to capture currency features, while
image processing techniques are employed to extract and analyze distinct features of both genuine
and fake banknotes. The objective is to create a robust system that can efficiently differentiate
between real and counterfeit currency notes, addressing the challenges posed by the proliferation of
counterfeit money in India's cash transactions[3].
 Detection of Fake Currency using Machine Learning
The system discussed in the report focuses on the detection of counterfeit currency using
machine learning and image processing, with a particular emphasis on Convolutional Neural
Networks (CNNs). It outlines a comprehensive methodology, including data collection, preprocessing,
feature extraction, and image classification, utilizing a pre-trained CNN model. The system also
addresses the potential for real-time currency detection in live video streams. While the report
highlights the importance of dataset quality and balance, it calls for further testing to evaluate system
efficiency and presents an accuracy graph from experimental results[4].
 Fake Currency Detection App
The fake currency detection app described in the study utilizes a sample dataset comprising
Indian currency notes. The primary method employed for counterfeit detection within the app is
Convolutional Neural Network (CNN) algorithm. The application is designed to be compatible with
smartphones having a storage capacity of 25GB and above. While the app demonstrated its capability
to identify counterfeit currency, it was noted that the user interface design was somewhat suboptimal,

16
leading to a somewhat clumsy user experience. Nevertheless, the app's fundamental functionality in
detecting counterfeit notes remains a noteworthy achievement in the realm of currency security
technology[5].
 APPROACH FOR HYBRID FAKE CURRENCY DETECTION SYSTEM
The hybrid fake currency detection system presented in this study employs a
comprehensive approach, primarily leveraging image processing techniques to achieve a wide range
of processing capabilities. The system has been meticulously trained to ensure high levels of accuracy
in detecting counterfeit currency. Through rigorous testing and evaluation, the developed web
application has demonstrated its ability to produce reliable and precise results. Notably, the security
measures implemented within the application make it resistant to unauthorized access, ensuring that
only authorized users can utilize its capabilities effectively. This system's multifaceted approach to
fake currency detection, coupled with its robust security features, positions it as a promising tool for
safeguarding financial transactions and mitigating the risks associated with counterfeit currency
circulation[6].
 Fake currency detection using image processing
The research study involves the development and implementation of a system for detecting
counterfeit Indian currency notes through image processing techniques. This system employs a digital
camera to capture ultraviolet images of currency notes, extracts various features such as the security
thread, serial number, and watermark, and then calculates their intensities to determine the
authenticity of the note. The system has shown promising results in distinguishing between genuine
and fake currency notes, with potential applications in automated teller machines and other currency
handling machines[7].
 A Novel Approach for Detection of Counterfeit Indian Currency Notes Using Deep
Convolutional Neural Network
In the context of the research on "A Novel Approach for Detection of Counterfeit Indian
Currency Notes Using Deep Convolutional Neural Network," a system study involves a thorough
investigation into the current processes, technologies, and challenges related to the detection of
counterfeit currency in India. This study examines existing methods, data sources, and technologies
used in currency verification, identifies shortcomings, and aims to develop a more accurate and
efficient system for detecting counterfeit Indian currency notes by implementing a three-layered Deep
Convolutional Neural Network (Deep ConvNet). The study assesses the feasibility and effectiveness
of this new approach in addressing the issue of counterfeit currency in the Indian economy[8].
 Research on Fake Indian Currency Note Detection using Image Processing
The research paper on "Fake Indian Currency Note Detection using Image Processing"
conducts a comprehensive system study to address the growing issue of counterfeit currency notes in

17
India. The study employs various image processing techniques, including Structure Similarity Index
Metric (SSIM), Adaptive Histogram Equalization, Fast Discrete Wavelet Transform (FDWT), and
Gray Level Co-occurrence Matrix (GLCM), in conjunction with machine learning via Artificial
Neural Networks (ANN). Through this systematic approach, the paper aims to develop a robust
system capable of accurately detecting fake Indian currency notes and determining their respective
denominations, ultimately contributing to the mitigation of economic challenges posed by counterfeit
currency[9].

18
 3 DESIGN

3.1 SYSTEM REQUIREMENTS

3.1.1 Software Requirements:

Jupyter Notebook: Jupyter Notebook provides an interactive environment for developing and testing
image processing algorithms, making it a suitable platform to experiment with various techniques.

Python: Python is a versatile programming language commonly used for image processing tasks. It
offers a wide range of libraries and tools for image analysis and manipulation.

Tkinter: Tkinter is a Python library used for creating graphical user interfaces. It can be employed to
design an intuitive user interface for the currency note detection system, making it user-friendly and
accessible.

Numpy: Numpy is a fundamental library for numerical and array operations in Python. It plays a
crucial role in image processing by enabling efficient data handling and manipulation.

Matplotlib: Matplotlib is a powerful library for creating visualizations, including displaying

processed currency note images and analysis results, making it essential for presenting the outcome of
the detection process[1][6][11].

3.1.2 Hardware Requirements

4GB RAM: A minimum of 4GB RAM is recommended to ensure efficient processing of image data,
allowing for smooth execution of image analysis algorithms without significant performance
bottlenecks.

Processor: A capable processor, such as a multi-core CPU with good clock speed, is crucial for
handling the computational demands of image processing tasks. A higher-end processor can
accelerate the detection process.

Hard Disk: Sufficient hard disk storage space is required to store image datasets, processing libraries,
and the application itself. Adequate disk space is essential for handling the large image files and the

19
software environment[1][5][9]

3.2 UML
3.2.1 Use case:

They are usually used to illustrate the various actions taken by the application. They also show
the several users who can carry out these functions. Use-case diagrams fall under behaviour diagrams
due to their emphasis on the tasks carried out and the users (actors) who carry out these tasks.
The numerous tasks carried out by the application include:
o Select the note
o Upload image
o Processing the note
o Display result

Figure 3-1 Use Case Diagram

20
 3.2.2 Class Diagram
A class diagram is a foundational element in software engineering and the Unified Modeling
Language (UML), representing the static structure of a software system. It visualizes the system's
architecture by depicting classes, their attributes (data members), methods (functions), and the
relationships between classes. Class diagrams provide a blueprint for designing and understanding
object-oriented systems, facilitating communication among developers, architects, and stakeholders.
These diagrams play a crucial role in software development, serving as documentation for the
organization and responsibilities of classes within the application. Moreover, while not directly
generating executable code, class diagrams serve as a bridge between system design and
implementation, ensuring that the software aligns with the principles of object-oriented
programming languages[9].

Figure 3-2 class diagram

3.2.3 Activity Diagram

Activity diagram is another important diagram in UML to describe the dynamic aspects of the
system. Activity diagram is basically a flowchart to represent the flow from one activity to another
activity. The activity can be described as an operation of the system.
The control flow is drawn from one operation to another. This flow can be sequential,
branched, or concurrent. Activity diagrams deal with all type of flow control by using different
elements such as fork, join, etc[8][19].

21
 Figure 3-3 activity diagram

3.2.4 Statechart Diagram

A Statechart diagram, a key component of the Unified Modeling Language (UML), provides a
visual representation of a state machine. In software engineering and system modeling, a state
machine is a conceptual model that defines various states an object can exist in and how it
transitions between these states in response to external or internal events. Each state represents a
specific condition or mode of the object, and transitions depict the change from one state to another
triggered by events. Statechart diagrams are highly valuable for modeling complex systems,
particularly those with dynamic behaviours, as they offer a clear and intuitive way to illustrate how
an object's behaviour evolves in response to different inputs and conditions[12].

22
Figure 3.4State chart Diagram

23
 4 IMPLEMENTATION

4.1 MODULES
4.1.1 Dataset
Preparation of Dataset:
The first step is the preparation of a dataset containing images of different currency notes(both fake
and real) and images of different features of each of the currency notes.
The dataset will contain the following repositories: –
Sub- dataset for Rs. 500 currency notes
– Images of real notes
– Images of fake notes
– Multiple images of each security feature (template)
Sub- dataset of Rs. 2000 and Rs. 100 currency notes (Similar structure)
The various security features that we are considering are: (for Rs. 500 currency notes- Total 10
features)
– Rs. 500 in Devanagari and English script (2 features)
– Ashoka pillar Emblem (1 feature)
– RBI symbols in Hindi and English (2 features)
– 500 rupees written in Hindi (1 feature)
– RBI logo (1 feature) – Bleed Lines on Left and right side (2 features)

24
 4.1.2 Models

We conducted feature detection and matching using ORB:

After completing the necessary processing of the image, feature detection and matching is done using ORB.
Our dataset already contains the images of different security features present in a currency note (total 10).
Further, we have multiple images of varying brightness and resolutions corresponding to each security feature
(6 templates for each feature). Using the ORB algorithm, each security feature is detected in the test image. To
make the searching of the security feature (template image) easier and more accurate, a search area will be
defined on the test currency image where that template is most likely to be present. Then, ORB will be used to
detect the template in the test image and the result will be highlighted properly with a marker. This process
will be applied for every image of each security feature present in the dataset and every time the detected part
of the test image will be highlighted properly using proper markers[13][14].

4.1.3 Performance Metrics

Feature comparison using SSIM:
From the previous step, the part of the test currency image which matches with each of the templates will be
generated. In this method, the original template will be compared with the extracted feature and then a score
will be given for the similarity between the two images using SSIM.

SSIM (x, y) = (2µxµy + c1) (2σxy + c2)/ (µ2 x + µ2 y + c1)(σ 2 x + σ 2 y + c2)

The Structural Similarity Index (SSIM) is a scoring system that quantifies the image quality degradation that is
caused by processing such as data compression or by losses in data transmission. Basically, it looks for
similarity between two images. It is a part of skimage library and uses the above-mentioned formula to
calculate similarity. It returns a value between -1 and 1. Closer the SSIM to 1, higher is the similarity. So, for
every security feature, the SSIM value between each image of that security feature and the corresponding
extracted feature from the test image will be calculated. Then, the mean SSIM for each security feature is
calculated and stored[1].

25
 .

4.2 OVERVIEW TECHNOLOGY

Currency counterfeit detection relies on a combination of advanced technological
components to accurately identify counterfeit banknotes while ensuring a user-
friendly and efficient operation. The core technological elements can be categorized
image processing technqiues[11][17].

4.2.1 TECHNOLOGY DETAILS

IMAGE PROCESSING TECHNIQUES

ROLE IN THE SYSTEM:

Image processing techniques are the foundation of the counterfeit detection process,
allowing for the precise analysis of currency notes. This component encompasses
several critical stages:

Image Acquisition: Currency notes are captured using various devices such as
scanners, cameras, or specialized currency recognition machines. High-quality images
are essential for accurate analysis.

Preprocessing: Image preprocessing is essential to enhance the quality of the

currency note images. This includes noise reduction to eliminate unwanted artifacts
and converting images to grayscale for uniformity.

Image Segmentation: The segmentation process divides the image into distinct
regions by establishing a threshold value. This separation is vital for isolating
currency notes from the background, facilitating further analysis.

Feature Extraction: Relevant features are extracted from currency note images.
These features serve as discriminators, allowing the system to distinguish between
genuine and counterfeit notes effectively.

User Interface: A user-friendly interface is provided for operators and financial

institutions. This interface streamlines the process of inputting and processing
currency notes, ensuring ease of use and efficient operation[19].

26
4.2.2. TKINTER

ROLE in the System:

Tkinter plays a crucial role in creating the graphical user interface (GUI) for the
counterfeit detection system. Its significance lies in the following aspects:

Tkinter is a standard Python library that simplifies the development of interactive

and user-friendly desktop applications with graphical interfaces.

Being part of the Python standard library, Tkinter eliminates the need for additional
package installations, making it a convenient choice for cross-platform GUI
application development in Python[16][18][20].

4.2.3. PYTHON

ROLE IN THE SYSTEM:

Python serves as the backbone of the entire currency counterfeit detection system,
orchestrating its various components. Python's role encompasses the following:

Python libraries like OpenCV are utilized for image capture and processing,
leveraging its robust image processing capabilities.

Numpy, a fundamental Python library, provides support for numerical and array-
based computations, enabling efficient data manipulation and analysis.

Matplotlib, another Python library, facilitates the creation of visualizations, charts,

and graphs, ensuring data representation and visualization for effective decision-
making[5][8].

27
 5 TESTING
5.1 TEST CASES

Table 5.1-1 Test Cases

Test Case Test Case Expected Actual Post Conditions

Output Output (yes/no)
ID Description
TC_01 Rs-500 in Compare and Test data should
Devanagai predict meet model
and English requirements Yes
script

TC_02 Ashoka Compare and Test data should

Pillar predict meet model
Emblem requirements Yes

TC_ 03 RBI symbols Compare and Test data should

In Hindi and predict meet model
Yes
English requirements

TC_04 500 rupees Compare and Test data should

Written in predict meet model
Hindi requirements Yes

TC_05 RBI logo Compare and Test data should

predict meet model
requirements Yes

TC_06 Bleed lines on Compare and Test data should

Left and predict meet model
Right side requirements Yes

28
 6 RESULTS
This section summarizes the quantitative results obtained through comprehensive
testing and evaluation, providing insights into the predictive performance of image processing
technique in detection of fake or real note.
 ACCURACY OF MODEL
Table 6.1 Accuracy
The performance analysis of the proposed system was carried out using various images of currency
notes. As we had already created a dataset for both fake and real currency notes of denominations of
100, 500 and 2000, all the notes were tested and then the accuracy was calculated. For the sake of
calculating the accuracy, it was assumed that if the currency note passed at least 9 features out of 10
then the note is real otherwise it is fake. Testing of both real and fake notes was
done separately.
• For testing of real notes, 93 real notes for Rs.2000 and 100 real notes for Rs. 500 were considered,
out of which 192 of the total 200 notes gave the correct desired results.
Accuracy: 96%
Webapplication

Fig – 6.1 Website

f

29
 Figure 6-2 Inserting 500 rupee note

Figure 6-3 All test features are passed

30
 Figure 6-4 2000 rupee note sent for processing

Figure 6-5 All features are not passed

31
 7 CONCLUSION

In this paper, a robust fake currency detection model was developed to authenticate Indian currency
notes, particularly those of denominations 100, 500, and 2000. This model was implemented using
the powerful OpenCV image processing library in Python3. It focused on analyzing ten critical
features of the input currency notes through the application of three distinct algorithms. To make the
model user-friendly, an intuitive graphical user interface (GUI) was created using the Tkinter library,
allowing users to easily upload and process images of currency notes from their systems.

The results of this model were promising, demonstrating its efficiency and accuracy. When
processing an input image, it took a mere five seconds, making it swift and practical for real-world
use. In terms of performance, the model showcased a commendable 93% accuracy in correctly
identifying genuine currency notes and an impressive 96% accuracy in detecting counterfeit
currency. This indicates that the model holds significant potential for aiding in the authentication of
Indian currency notes, contributing to increased security and trust in financial transactions. Overall,
this research presents a valuable tool in the ongoing efforts to combat counterfeit currency issues.

32
 8 FUTURE SCOPE

The future scope for the fake currency detection system is promising and offers numerous
opportunities for enhancement and expansion:
Additional Features: One avenue for future work involves adding more features to the system for
even more robust counterfeit detection. Researchers can explore and incorporate advanced security
elements present in currency notes to further improve the model's accuracy.
Expanding to Other Denominations: While the current system focuses on 100, 500, and 2000 rupee
notes, it can be extended to cover the remaining denominations as well. This would make the system
more comprehensive and beneficial for users dealing with a variety of currency notes.
Integration with Mobile Apps: Developing a mobile application that integrates with the counterfeit
detection system can significantly enhance its accessibility. Users can simply capture images of
currency notes using their smartphones and receive instant authenticity verification, making it user-
friendly and widely accessible.
Accessibility Features: To cater to a wider audience, incorporating accessibility features such as text-
to-speech or tactile feedback for visually impaired users can make the system more inclusive.
Security Updates: Regular updates and security patches should be implemented to stay ahead of
counterfeiters who may develop new techniques to produce fake currency.
Collaboration with Authorities: Collaborating with law enforcement agencies and central banks can
help improve the system's effectiveness and provide valuable insights into emerging counterfeit
trends and techniques

33
 BIBILOGRAPHY

[1] Kamble, K., Bhansali, A., Satalgaonkar, P., & Alagundgi, S. (2019, December).
Counterfeit currency detection using deep convolutional neural network. In 2019 IEEE
Pune Section International Conference (PuneCon) (pp. 1-4). IEEE.

[2] Veling, S., SAWAL, J., BANDEKAR, S., PATIL, T., & SAWANT, A. (2019). Fake Indian
Currency Recognition System by using MATLAB. International Journal for Research in
Applied Science & Engineering Technology, Sindhudurg, Maharashtra.

[3] Shiby, A., Francis, F., Jose, P., & Augustine, R. (2021). Detection of Fake Currency
Using Image Processing. International Journal for Research in Applied Science and
Engineering Technology.

[4] Patil, D. P., Varma, G., Poojary, S., Sawant, S., & Sharma, A. Counterfeit Currency
Detection based on AI.

[5] Gaikwad, A., Dudka, S., Shah, D. G., Mitkar, S., & Kumbhar, P. FAKE CURRENCY
DETECTION WEB APPLICATION.

[6] Kore, S. L., Mishra, D., Bhiogade, I., & Jituri, D. FAKE CURRENCY DETECTION USING
RECURRENT NEURAL NETWORKS (RNN).

[7] Prof Chetan More, Monu Kumar, Rupesh Chandra, Raushan Singh, “Fake currency Detection
using Basic Python Programming and Web Framework” IRJET International Research Journal of
Engineering and Technology, Volume: 07 Issue: 04 | Apr 2020 ISSN: 2395- 0056

[8] Vivek Sharan, Amandeep Kaur,” Detection of Counterfeit Indian Currency Note Using Image
Processing” International Journal of Engineering and Advanced Technology (IJEAT), Volume.09,
Issue:01, ISSN: 2249-8958 (October 2019)

[9] Aakash S Patel, “Indian Paper currency detection” International Journal for Scientific Research &
Development (IJSRD), Vol. 7, Issue 06, ISSN: 2321-0613 (June 2019)

[10] Archana M Kalpitha C P, Prajwal S K, Pratiksha N,” Identification of fake notes and
denomination recognition” International Journal for Research in Applied Science & Engineering
Technology (IJRASET), Volume. 6, Issue V, ISSN: 2321-9653, (May 2018)

[11] S. Atchaya, K. Harini, G. Kaviarasi, B. Swathi, “Fake currency detection using Image
processing”, International Journal of Trend in Research and Development (IJTRD), ISSN: 2394-9333
(2017).

[12] S. R. Darade and G. Gidveer, “Automatic recognition of fake indian currency note,” in 2016
international conference on Electrical Power and Energy Systems (ICEPES). IEEE, 2016, pp. 290–
294.

34
 [13] B. P. Yadav, C. Patil, R. Karhe, and P. Patil, “An automatic recognition of fake indian paper
currency note using matlab,” Int. J. Eng. Sci. Innov. Technol, vol. 3, pp. 560–566, 2014.

[14] A. Zarin and J. Uddin, “A hybrid fake banknote detection model using ocr, face recognition and
hough features,” in 2019 Cybersecurity and Cyberforensics Conference (CCC). IEEE, 2019, pp. 91–
95.

[15] M. S. Veling, M. J. P. Sawal, M. S. A. Bandekar, M. T. C. Patil, and M. A. L. Sawant, “Fake

indian currency recognition system by using matlab.”

[16] F. A. B, P. Mangayarkarasi, Akhilendu, A. A. S, and M. K, “Fake indian currency note

recognition,” vol. 7, pp. 4766–4770, 2020. [Online]. Available:
https://www.irjet.net/archives/V7/i5/IRJET-V7I5915.pdf

[17] M.Gogoi, S. E. Ali and S. Mukherjee 2015 Automatic Indian currency denomination recognition
system based on artificial neural network 2nd International Conference on Signal Processing and
Integrated Networks (SPIN) IEEE 553-558.

[18] S. Murthy, J. Kuru Mathur and B. R. Reddy 2016 Design and implementation of paper currency
recognition with counterfeit detection Online International Conference on Green Engineering and
Technologies (IC-GET) Coimbatore 1-6.

[19] S. Singh, S. Choudhury, K. Vishal and C. V. Jawahar 2014 Currency Recognition on Mobile
Phones 22nd International Conference on Pattern Recognition Stockholm 2661-2666.

[20] Alekhya, D., Prabha, G. D. S., & Rao, G. V. D. (2014). Fake currency detection using
image processing and other standard methods. International Journal of Research in
Computer and Communication Technology, 3(1), 128-131.

35