CNN based Counterfeit Indian Currency Recognition Using Generative Adversarial Network

In today's world scenario, paper currency is economical in the sense that its face value is greater than intrinsic value. It is also more elastic and stable, paper currency can be counted quickly, it is easy to move and safe to store. These all are the main reasons because of which counterfeit currency recognition is crucial. Fake currency cannot be identified by human vision and due to this recognition of forged currency notes has become crucial problem because counterfeiters are using new and improved methods. The methods currently existing to determine whether the notes are real cannot be accessed by the common people and are also complex hardware based methods. There are no applications or devices available through which fake currencies can be detected and identified easily by common people. The main purpose of the project is to identify Indian paper currency with a new methodical approach using Generative Adversarial Networks(GAN). In this system, the Indian currency note features would be primarily extracted using Convolutional Neural Networks (CNNs).The processed image data are then fed to a Generative Adversarial Network which helps to classify the currency as either real or fake. GAN consists of two main modules - Generator and Discriminator. The Generator generates fake currency images and the Discriminator identifies and labels the real and fake images.