Research And Design Of Intelligent Recognition System For RMB Banknotes

Current market daily circulation of paper money is a very large number of and within the financial industry notes sorting finishing work is heavy,how fast and accurate for banknotes are clearing in the financial industry in the existing difficulties and problems,banknote sorting algorithm can make the tedious work of paper currency sorting becomes easy,fast and reliable,work efficiency so as to effectively improve the financial industry,starting from this point,the research and design of RMB notes intelligent recognition system has important meaning and value.In this thesis,the problem of outlier detection and elimination in the intelligent recognition of RMB notes and the current research status are reviewed,and the limitations of the research are analyzed and pointed out.According to the domestic general yuan,using digital image processing and pattern recognition technology,a set of thesis currency recognition algorithm design and implementation,including sorting,sorting,oriented towards the value of old and new types of sorting,sorting and grading of incomplete cleaning.The specific research contents are as follows:(1)at the same time sorting algorithm designed in this thesis does not affect the accuracy of the banknote sorting complexity is low,greatly reducing the time-consuming procedures,can meet the real-time banknote sorting system is fit for the.Before the thesis currency recognition,preprocessing operation on the banknote image,including thesis edge detection,feature extraction,notes notes image enhancement and image binarization,the clever use of the algorithm to extract the features of regional currency,in the feature extraction and the target area has been rotating.(2)the optimization design method of detection and removing abnormal sequence notes in the edge points,and accuracy of the least squares linear fitting and linear fitting method in the presence of outliers in the notes fitting edge line for improvement of research.Image processing is the most important value of two image after combining the two mainstream binarization algorithm of Otsu and Bernsen,designed a better binarization algorithm can reduce the complexity of the algorithm and reduce the running time of the program,is conducive to the subsequent recognition of banknote image classification.This thesis has finished the RMB notes intelligent recognition system of paper currency image pre processing algorithm,currency sorter recognition algorithm,through the simulation experiment reached expected requirements.