Research On The Recognition Method Of Euro Banknotes

Paper currency is one of the mainstream tools in the world trade.With the further promotion of economic globalization,the amount of trade among countries is increasing,and the demand for paper currency is increasing.When checking paper currency manually,the efficiency is low and the error rate is high,so the paper currency sorter appears.The core technology of paper currency sorter is to recognize the denomination of paper currency,which replaces the manual work of paper currency sorting,improves the efficiency and accuracy of paper currency sorting,and plays an important role in the field of banking and other financial services.Due to the gradual electronization of the Chinese Renminbi,as a major producer,China has developed banknote clearing products mainly for export and sales all over the world.Since the Russian-Ukrainian war,the position of the Euro in the international monetary system has been declining,but the proportion of global payments has risen to a new stage with 36.12%,which shows that the Euro is still recognized internationally.According to foreign trade data,in the two months before 2022,the European Union anti-super-ASEAN became China’s largest trading partner.All kinds of signs indicate that the research and export of Euro Banknote Cleaner is imperative.In addition,the Euro Banknote Cleaner designed and produced in China has a lower cost and a better price competitive advantage than similar products designed in developed countries.For the recognition of Euro currency,this paper puts forward a different method from the previous manual extraction of the features of Euro currency.The method of deep learning is used to recognize Euro currency face-to-face and improve the recognition accuracy by data enhancement and attention mechanism in residual network.Specifically,first,in view of the problem that there is no public image data collection of Euro notes,this paper creates an image collection of Euro notes by collecting images of Euro notes,and obtains 8 types of over 3000 Euro notes pictures as a dataset for Euro notes recognition.Secondly,by comparing the complexity and applicability of network models,this paper makes experiments on common network models such as AlexNet,VGNet,GoogleNet,ResNetl 8,ResNet34 and ResNet50,and finally selects ResNet50 as the basic network.In order to further improve the accuracy of the recognition of Euro currency,this paper presents a Euro currency recognition model based on the improved ResNet50 neural network.The model is optimized and improved on the ResNet50 basic network model.The attention mechanism module is introduced.By using the attention mechanism,areas with distinct changes in different state characteristics can be captured and focused,which not only keeps the parameters of the model,but also improves the classification accuracy.Optimized network performance.The results show that the recognition accuracy of Euro notes can reach 99.999%,which ensures high accuracy and improves the recognition speed.It can be seen that compared with the traditional algorithm,the method proposed in this paper improves the overall recognition rate in face and face recognition,and the algorithm has strong robustness,can meet the needs of different directions of money entry,and is less affected by external factors.