The Design And Implementation Of The Recognition For Face Value And Serial Number Of Paper Currency

Nowadays, with the development and progress of society, many industries have realized the automation and informationization. At the same time, comprehensive automation will also be the inevitable trend for the banking business. The application of intelligent Currency Sorter is a reflection of the automatic banking business. In this paper, the recognition of face value and Serial Number of paper currency which are the core technology of Currency Sorter is discussed deeply.In the research of preprocessing for paper currency image, according to the problem of image segmentation, a method of fitting a straight-line for the edge of paper currency is proposed, which successfully solves the false detection of the paper currency edge. As for the deformation problems of paper currency image, this paper puts forward a deformation correction method based on image transformation model, this method solves various linear deformation problems via the mapping relationship between the original image and standard template,In the face value identification of paper currency, this paper proposes an identification algorithm based on blocking LBP histogram. This algorithm can achieve a recognition rate of 100% for the image acquired by scanner, but a much lower recognition rate for the image acquired by Currency Sorter. In order to solve this problem, this paper proposes another identification method based on phase correlation, the recognition rate can also reach 100%.This paper adopts connected domain method for the location of the Serial Number, this method makes use of the rules that the rectangles of the character connected domains satisfy and detects the location parameter of the Serial Number. As for the influence caused by background texture in the process of segmentation the Serial Number characters, this paper preprocesses the serial number image through gray stretch. In this way, contrast ratio between background and character is enhanced, thus the negative influence to characters segment is eliminated.