Currency Recognition Using Image Processing

Abstract

The circulation of counterfeit currency has become a growing concern in modern financial systems, especially in developing countries where cash transactions remain dominant. Traditional methods for detecting fake banknotes mainly rely on manual inspection or dedicated hardware devices such as ultraviolet detectors. These methods are often costly and not easily accessible to common users. In this project, an automated system for Indian currency recognition and counterfeit detection based on digital image processing techniques is presented. The proposed approach uses image acquisition, preprocessing, segmentation, feature extraction and classification to identify and verify currency notes. Statistical and edge-based features are extracted from the currency images and compared with reference feature sets stored in a database. The system is implemented using MATLAB and is evaluated on different denominations of Indian currency. Experimental results indicate that the proposed method is capable of recognizing genuine notes and detecting suspicious notes with satisfactory accuracy, making it suitable for low-cost and small-scale deployment.