Design And Implementation Of Commercial Bank Form Recognition Based On Image Processing Technology

In order to expand the credit card business,a large number of paper bank forms for credit card application are generated every day.To realize the informatization of bank,it is necessary to input the information on paper bank form into the system.At present,the manual input model has low work efficiency and high labor intensity,so it is an urgent problem for people to design a system that can automatically import the text information in the paper bank form into the computer.This paper takes the credit card application form of a bank as an example and designs a method of bank form recognition based on image processing technology.In this paper,a series of image preprocessing,including image correction,image denoising,binary processing and layout analysis based on projection method,are carried out.Specifically,the form coordinates and the standard model are used to obtain the slant Angle of the scanned image,and the slant Angle is reversed to achieve correction;median filter is used to remove the "salt and pepper" noise mixed with the scanned image;an iterative Otsu method is designed to perform binarization processing on the denoised image;the layout of binary image is analyzed based on projection method.Based on the edge detection method,a positioning method combining image morphological processing,contour detection,external rectangle and template matching was designed.In order to make the final detection result more accurate,the post-processing process is added to calibrate the automatic identification result after the bank form automatic identification is completed.Finally,the whole implementation of the system is described.To prove the reliability of the system,a series of test experiments are carried out,and the text area positioning and text automatic recognition test experiments are carried out in detail.