Design And Implement On Multiple-Modes Banknote Counterfeit System

Technique of Banknote counterfeit has always been connected with domestic financial system and economoy field. Because there lies many characteristics in a single banknote to counter fake money, and images are illustrated in low resolution, the problem of how to recognize and identify banknote has been the core issue.In this paper we find banknote will expose unique characters in different modes. So we use the IR, UV image scanned from the banknote sorting machine via CIS and magnetic signal to design a multiple-mode banknote validator. We introduce the way to locate IR and UV characters in image and way to extract effective information to be used in validator and introduce distance-matching algorithm to classify different kinds of banknote, which proves to be valid. On the basis of traditional ways of identifying magnetic signal, we introduce a way to visualize magnetic signal to make it more readable and understandable. We introduce three ways to calculate the similarity between two modes as follow: positive & negative model, absolute distance and vector similarity calculating. In validator, we use the image processing way instead of digital signal processing method to process and identify magnetic signal, which proves to be available.In designing multiple-modes validator system strategy, we introduce a weight value to every result given by different module and designed the final results by combining the result returned from image pre-processing module, IR validator module, UV validator module and magnetic signal module.The final result consists of three parts and this method assembles various counterfeit techniques and bring huge convenience to the finance field.