Currency Recognition Using Image Processing and Minimum Distance Classifier Technique

Abstract

As a result of the great technological advances over the past few years in duplicating and scanning, counterfeiting problems have become more and more serious. Fake banknotes have become so deeply embedded in the Indian economy that even bank branches and ATMs are disbursing counterfeit currency. Estimation of fake currency in Indian economy is about 10-20 percent of total notes in circulation. From petrol stations to the local vegetable vendors, everybody is wary of accepting banknotes in denomination of 500 and 1000. It is difficult for people to recognize and detect the notes. To solve these problems a currency recognition system is implemented to reduce human power to automatically recognize currency without human supervision. The software interface that has been proposed in this paper is to recognize the Indian currency notes and authenticate to certain extent. Digital image processing is one of the most common and effective techniques used to distinguish counterfeit banknotes from genuine ones. This paper introduces a recognition and detection method for Indian currency using Image Processing. It is shown that Indian currencies can be classified based on a set of unique non discriminating features such as dominant color, dimension, latent image and Identification Mark mentioned in RBI guidelines. Firstly, the aspect ratio and the dominant color of the note are extracted. After this the segmentation of the ID mark and latent image is done. Segmented features are then processed and classified using Minimum Distance Classifier.