Research On Recognition Algorithm Of RMB Paper Currency Number

China is a country with a large population,RBM has a large circulation,with the rapiddevelopment of market economy,a lot of social problems has emerged,we can find many ofvicious crime occurred for cash related, for example, robbing a bank, kidnapping, manufacturingforged note.To realize the fast recognition number using OCR(Optical CharacterRecognition),and complete paper currency tracking. So as to solve the above problems. Thedomestic and foreign relevant institutions have done a lot of work and study,but there still existsome deficiencies.RMB number image processing algorithm and recognition algorithm are key RMB ofnumber automatic recognition. In fifth edition one hundred yuan RMB as the research object,And the research of number image preprocessing included Fast Filtering Algorithm,RMBNumber Location,Image Binarization,Character Segment,Character Size Normalization and soon. Presented a number recognition algorithm based on multiple mutual compensation structurecharacteristics used in RMB paper currency number recognition.And the recognition algorithmand the algorithm flow chart were given.An ARM_based embedded system was used to verify programs. Some schemes weresuggested for solving differences in Characters’ fonts and character interference issues whichhave arose in the experiment.The statistical results of accuracy of recognition programs showedthe algorithm has high precision and meets the practical needs.