FAKE CURRENCY DETECTION
USING K-NN ALGORITHM
PRESENTED BY:-
ENJOY
TABLE OF CONTENT
• Introduction
• Techniques to detect currency counterfeit
• Performance analysis
• K-NN Algorithm
• How K-NN is used in detecting counterfeiting
• Disadvantages of using K-NN
• Conclusion
• Future Work
INTRODUCTION
• Currency counterfeiting is always been a challenging term for financial
system of any country. The problem of counterfeiting majorly affects the
economical as well as financial growth of a country, the race is going on
between the counterfeiters and the banks. To resolve the issue various
researchers came across with variety of techniques and proposed solutions
mostly focused on Machine learning and Image processing areas
TECHNIQUES TO DETECT COUNTERFEIT
Currency counterfeit can be detected by different techniques
• By checking watermarks , ultra violet waves , various image processing
techniques like canny edge detection , sobel operator, various
machine learning algorithms.
• Nowadays researchers mainly focus on Machine learning and Image
processing areas.
Here are performance analysis of some ML
algorithms:-
Out of these three ML algorithms K-NN performed with highest
accuracy rate .
K-NN ALGORITHM
• K-NN uses the concept of feature similarity to find out the new data point
values i.e., the value assigned to the new data point is based on the
matching of its value to the points of training set.
• K-NN groups the data into coherent clusters or subsets and classifies the
newly inputted data based on its similarity with previously trained data.
The input is assigned to the class with which it shares the most nearest
neighbors
Working of K-NN algorithm:-
• Select the number K of the neighbors
• Calculate the Euclidean distance of K number of neighbors with all training data
points using the formula
• Take the K nearest neighbors as per the calculated Euclidean distance
• Among these k neighbors, count the number of the data points
in each category.
• Assign the new data points to that category for which the number of the
neighbor is maximum.
KNN (K-NEAREST NEIGHBOR) METHOD IN
CURRENCY COUNTERFEIT:-
The input data set was made by collecting high-quality images of both
genuine and counterfeit currency using an industrial camera. On these
pictures, Wavelet Transform was used to extract features from the
gathered images. The attributes gathered after the Wavelet
Transformation were as follows-
• Variance
• Skewness
• Kurtosis
• Entropy
• Class of the currency
- Data set is obtained by plotting various attributes.
- Preprocess and normalize the data set.
data needs to be scaled so that the model is not biassed towards a
particular feature
- Split the data using K-fold cross validation technique.
- Training the model using the algorithm.
- Calculate the performance measures.
DIS ADVANTAGES OF USING K-NN ALGORITHM
• Does not work well with large dataset
• Does not work well with high dimensionality
• Sensitive to noisy and missing data
• Feature Scaling
CONCLUSION:-
• K-NN is one of the best machine learning algorithm for detection of
currency counterfeit.
• However it doesn’t work well for large datasets
• In this case the algorithm has to deal with image processing and it doesn’t
goes well with image processing because of its sensitivity to noisy and
missing data.
FUTURE WORK
• For large data set available, deep learning algorithms like Convolutional
Neural Networks or CNN can be applied which have high accuracy in
image processing scenarios. Furthermore, by using CNN the project can
directly analyze images as input, and wavelet transformation will not be
required. This can make the system more convenient and user friendly to
use.
Reference:-
• https://doi.org/10.1109/ICICCS51141.2021.9432274
• https://doi.org/10.1109/ICCS45141.2019.9065747
• https://doi:10.1088/1757-899X/1073/1/012029
THANK YOU!

Fake currency detection using knn algorithm.pptx

  • 1.
    FAKE CURRENCY DETECTION USINGK-NN ALGORITHM PRESENTED BY:- ENJOY
  • 2.
    TABLE OF CONTENT •Introduction • Techniques to detect currency counterfeit • Performance analysis • K-NN Algorithm • How K-NN is used in detecting counterfeiting • Disadvantages of using K-NN • Conclusion • Future Work
  • 3.
    INTRODUCTION • Currency counterfeitingis always been a challenging term for financial system of any country. The problem of counterfeiting majorly affects the economical as well as financial growth of a country, the race is going on between the counterfeiters and the banks. To resolve the issue various researchers came across with variety of techniques and proposed solutions mostly focused on Machine learning and Image processing areas
  • 4.
    TECHNIQUES TO DETECTCOUNTERFEIT Currency counterfeit can be detected by different techniques • By checking watermarks , ultra violet waves , various image processing techniques like canny edge detection , sobel operator, various machine learning algorithms. • Nowadays researchers mainly focus on Machine learning and Image processing areas.
  • 5.
    Here are performanceanalysis of some ML algorithms:- Out of these three ML algorithms K-NN performed with highest accuracy rate .
  • 6.
    K-NN ALGORITHM • K-NNuses the concept of feature similarity to find out the new data point values i.e., the value assigned to the new data point is based on the matching of its value to the points of training set. • K-NN groups the data into coherent clusters or subsets and classifies the newly inputted data based on its similarity with previously trained data. The input is assigned to the class with which it shares the most nearest neighbors
  • 7.
    Working of K-NNalgorithm:- • Select the number K of the neighbors • Calculate the Euclidean distance of K number of neighbors with all training data points using the formula • Take the K nearest neighbors as per the calculated Euclidean distance • Among these k neighbors, count the number of the data points in each category. • Assign the new data points to that category for which the number of the neighbor is maximum.
  • 9.
    KNN (K-NEAREST NEIGHBOR)METHOD IN CURRENCY COUNTERFEIT:-
  • 10.
    The input dataset was made by collecting high-quality images of both genuine and counterfeit currency using an industrial camera. On these pictures, Wavelet Transform was used to extract features from the gathered images. The attributes gathered after the Wavelet Transformation were as follows- • Variance • Skewness • Kurtosis • Entropy • Class of the currency
  • 11.
    - Data setis obtained by plotting various attributes. - Preprocess and normalize the data set. data needs to be scaled so that the model is not biassed towards a particular feature - Split the data using K-fold cross validation technique. - Training the model using the algorithm. - Calculate the performance measures.
  • 12.
    DIS ADVANTAGES OFUSING K-NN ALGORITHM • Does not work well with large dataset • Does not work well with high dimensionality • Sensitive to noisy and missing data • Feature Scaling
  • 13.
    CONCLUSION:- • K-NN isone of the best machine learning algorithm for detection of currency counterfeit. • However it doesn’t work well for large datasets • In this case the algorithm has to deal with image processing and it doesn’t goes well with image processing because of its sensitivity to noisy and missing data.
  • 14.
    FUTURE WORK • Forlarge data set available, deep learning algorithms like Convolutional Neural Networks or CNN can be applied which have high accuracy in image processing scenarios. Furthermore, by using CNN the project can directly analyze images as input, and wavelet transformation will not be required. This can make the system more convenient and user friendly to use.
  • 15.
  • 16.

Editor's Notes