Research On Infrared Feature Authentication And Counterfeiting Algorithm Based On PCA + SVM Paper Currency

Banknote recognition is an important issue in today's society and an important area of modern popular deep learning and artificial intelligence.At present,the identification of banknotes has been studied for a long time.In recent years,along with economic development,the authenticity of banknotes has affected our daily life.Most of the enterprises have implemented digital and intelligent banknotes The extension of the realization of the integration process.However,there is a certain accuracy of the sorter,susceptible to external factors and anti-pollution ability.This article carries on the research and analysis of the authenticity verification after the paper currency pretreatment,and then optimizes the paper currency's infrared characteristic and the true and the false currency classification.The previous banknote processing algorithm optimization,has been very good recognition.Finally,simulation test is carried out,and the result of experiment verification is analyzed.The paper currency used in this article is the fifth version of the RMB counterfeit sample image,before carrying out the identification of banknotes,we should do a good job on banknotes pretreatment.In order to improve the banknote recognition rate,Gabor filter processing is performed on the enhancement of the banknote after the preprocessing of the banknote to obtain the processed banknote image.Then the banknote infrared identification work is carried out.In this paper,PCA algorithm is used to identify the banknotes,the dimensionality reduction of the banknotes is carried out,the Euclidean distance is used as the calculation of the distance between the banknotes,The maximum distance as the true and false currency judgment threshold,the recognition rate of banknotes has increased,but the miscarriage of justice is not good.In view of such situation,in the following paper,the combination of PCA + SVM algorithm is used to optimize the kernel function and penalty factor of SVM.Finally,the characteristic amount of the banknotes to be processed and the processing time are reduced.More importantly,the recognition rate of the banknotes after the optimization of the SVM parameters is improved.Meanwhile,the misclassification rate of the counterfeit currency is greatly reduced,and in a certain sense Meet the requirements of the market,China's paper currency research has a good reference and reference value.