Fake Indian Currency Detection System

Abstract

This project introduces a new method of counterfeit Indian currency detection based on machine learning. It integrates image processing methods with sound classification algorithms, and the system examines digital images of banknotes to detect inconsistencies that point to counterfeiting. The method includes image pre-processing, feature extraction. The performance of the proposed method is tested using stringent experimentation on a real and spurious Indian banknote dataset and shows promise of being useful for real-world applications in the prevention of currency forgery.

Introduction

Overview

Currency counterfeiting poses severe economic risks globally, especially in India. Traditional detection methods—like visual inspection and UV light—are prone to human error, time-consuming, and less effective against advanced forgeries. This project proposes an automated, accurate, and scalable solution using machine learning (ML) and image processing to detect counterfeit Indian banknotes.

Key Components of the System

1. Basic Concept

The system learns to distinguish genuine and fake banknotes by analyzing their visual features using ML algorithms.

2. Workflow

• Data Acquisition: Collect images of both real and counterfeit notes across various denominations.

• Image Preprocessing: Resize, enhance, and denoise images for consistency and clarity.

• Feature Extraction: Analyze texture, edges, colors, and embedded security elements (e.g., watermark, thread).

• Model Training: Use algorithms like SVM, CNN, and Random Forests to classify notes.

• Testing & Evaluation: Validate the model’s accuracy using metrics like precision and recall.

• Deployment: Implement the model in real-world tools (e.g., mobile apps, cash counting machines).

Literature Review Highlights

• ML models (SVM, CNN) outperform traditional methods in accuracy.

• Assistive apps (like Sahayaka) help the visually impaired detect fake notes.

• Combining image processing and deep learning boosts detection capabilities.

Limitations of Traditional Methods

• Rely on manual inspection (watermarks, threads, microtext).

• Vulnerable to human error and inefficiency.

• Not scalable for high-volume transactions.

• Struggle with sophisticated fakes.

Proposed Architecture & Modules

• Image Acquisition: Captures clear, consistent note images.

• Preprocessing: Enhances image quality.

• Feature Extraction: Identifies distinct visual and security elements.

• ML Classification: Classifies notes using trained models.

• Output Module: Displays and logs results with confidence levels.

• Database: Stores images, features, and results for continual learning and auditing.

Advantages of the Proposed System

• Higher accuracy and objectivity vs. human checks.

• Faster, automated processing of large volumes.

• Detects subtle counterfeit features.

• Can adapt to new forgery techniques through retraining.

• Reduces dependency on human expertise.

• Scalable, cost-effective, and easy to integrate into ATMs or bank systems.

• Provides data logging for audits and trend analysis.

• Ensures consistent performance without fatigue.

Conclusion

The project created a machine learning-based system for the detection of counterfeit Indian currency with promising results in the identification of counterfeit banknotes. The performance of the system, however, relies on the quality and diversity of the training dataset. Future research will include increasing the dataset to cover more types of counterfeits, investigating more sophisticated feature extraction methods, and optimizing the machine learning algorithms for higher accuracy and resistance. In addition, investigating real-time deployment and integration with current currency processing equipment is an essential next step.

References

[1] Sanyal, S., Kaushik, A., & Gandhi, R. \"Spurious Currency Detection using Machine Learning Techniques.\" 2024 International Conference on Automation and Computation (AUTOCOM) [2] P. S, S. C, V. Padmapriya, and S. Uma. \"Sahayaka: A Fake Currency Detector Application for Visually Impaired Individuals.\" 2024 IEEE International Conference for Women in Innovation, Technology & Entrepreneurship (ICWITE) [3] VivekSharan, AmandeepKaur, Parvinder Singh. \"Identification of Counterfeit Indian Currency Note using Image Processing and Machine Learning Classifiers.\" Proceedings of the Third International Conference on Artificial Intelligence and Smart Energy (ICAIS 2023) [4] Kara, S. T., Loya, S., Raju, S. S., Vanteru, N., &Rajulapati, B. \"Detection of Fake Indian Currency Using Deep Convolutional Neural Network.\" 2023 IEEE 3rd Mysore Sub Section International Conference (MysuruCon) [5] Shinde, S., Wadhwa, L., Naik, S., Kudale, R., Sherje, N., &Mohnani, K. \"Fake Currency Detection using Image Processing.\" 2023 7th International Conference on Computing, Communication, Control and Automation (ICCUBEA) [6] Kumar, C. P., Yadav, M. G., Praneetha, K., Rushikesh, M., & Shreya, R. R. \"Classification and Detection of Banknotes using Machine Learning.\" 2023 5th International Conference on Inventive Research in Computing Applications (ICIRCA)

Copyright

Copyright © 2025 Prof. P. S. Purohit, Sanika Dattatraya Bhosle, Ashish Hemant Ghatule, Tejaswini Vijay Kolekar, Ritesh Arvind Kataktalware, Vishal Ramkrishna Patil. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.