Detection of Fake Indian Currency Using Deep Convolutional Neural Network

In the highly advanced technological world of today, counterfeit currency poses a significant threat, given that the face value of paper currency is much higher than its intrinsic value, and its elasticity and it is stable, which make it simple to transfer and secure to store. Despite the complexity of current hardware-based methods for detecting fake notes, the main problem is that these methods are not easily accessible to Ordinary individuals. This paper proposes a method of detecting fake currency by a hybrid approach using CNN and Resnet model to produce accurate results. This system utilizes convolutional neural networks (CNN) and residual networks (ResNet) to detect fake currency utilizing the width, colors, and serial numbers. The system involves stages, like image pre-processing, feature extraction, and identification. In the first stage, the scanned currency images are preprocessed to enhance the quality of the images and reduce noise. Then, features are extracted from the preprocessed images using CNN and ResNet architectures. The proposed automatic system for identifying Indian currency notes and checking whether they are real or fake uses CNN techniques. This system is very useful in banking and other fields, given the rise of counterfeit currency notes in India. Finally, the paper introduces a Deep Learning approach using a convolution neural network (CNN) model that can identify counterfeit notes on handy devices like smartphones and tablets, with a testing accuracy of about 98.3%.