Detection of Fake Currency Using Machine Learning

The project aims to address the escalating threat of counterfeit currency by proposing and implementing an innovative fake currency detection system. Counterfeiting poses a significant risk to both individuals and the national economy. Traditional detection methods are limited to banks and corporate offices, leaving ordinary citizens and small businesses vulnerable. In response, this project focuses on the security features of Indian currency notes, leveraging advanced image processing and computer vision techniques to develop a software-based authentication system. Implemented using Python in a Jupyter Notebook environment, the system meticulously analyzes key features, such as bleed lines, security threads, latent images, watermarks, and more, specific to denominations of 500 and 2000 rupees. The system incorporates three main algorithms to validate currency notes, ensuring a comprehensive examination. The first algorithm employs ORB detection and SSIM for feature extraction and comparison. The second authenticates bleed lines, while the third verifies the number panel of currency notes. The automated system provides a rapid and accurate means of detecting fake currency, replacing time-consuming manual methods. By creating a user-friendly interface, the project aims to empower individuals and businesses to safeguard against counterfeit currency effectively. The performance analysis indicates promising results, with an accuracy of 79% for genuine notes and 83% for counterfeit notes. This project introduces an accessible and efficient solution to enhance currency authentication for widespread use.