Indian Currency Detection using Image Recognition Technique

Abstract

The paper currency counterfeiting is a big problem for the world. Almost every country has been badly affected by this which has become a very acute problem. The main purpose behind this study is to recognize Indian paper currency with this hybrid approach which is portable and making an application used on the go. In this study, the Indian currency note features will be extracted and will be stored in MAT files and then these stored features will be matched with the input paper currency to recognize whether it is genuine or duplicate. With this system, easy to recognize the currency note anywhere, anytime. I have used the MATLAB image processing toolbox. The image processing is a way to improve the pictorial information of the image for the sake of machine or hardware perception. The currency notes will be recognized with the combination of both local binary patterns and principal component analysis. The LBP is significant progress in texture analysis and used for matching purpose. PCA is used for training purpose. Euclidian distance algorithm will be used for combining the metrics which has simple measure computations. Currency recognition has big challenges like watermark recognition, currency note resolution, dirty notes etc.