Currency Recognition and Counterfeit Detection Using Feature Extraction and AI

In today's digital economy, paper currency remains widely used, but advancements in technology have made counterfeiting easier, posing a serious threat. Manual counterfeit detection is time-consuming, highlighting the need for automated solutions. This work develops a system for fake currency detection and denomination recognition of Indian notes. Specifically, we worked with SIFT-FLANN algorithm for counterfeit detection and deep learning algorithms such as ResNet-18 for denomination classification. To enhance accuracy, we have implemented the latest YOLO version for counterfeit classification and denomination recognition. Moreover, we have demonstrated the interpretability of the ResNet-18 model for denomination recognition using LIME and Grad-Cam which would be beneficial in understanding the important factors behind the particular prediction. Additionally, our system features a user-friendly interface for verification and detection.