Method for extracting paper bank statement information by using sliding template technology

Abstract

The invention discloses a method for extracting paper bank statement information by using a sliding template technology, and the method comprises the steps: carrying out optical character recognition on a bank statement scanning copy to obtain the textbox position and content information of each page; making a template which is a rectangular box of multiple specifications, wherein an image area, recording the bank statement record, on the bank statement scanning copy is selected by the rectangular box; moving the relative position of the template and the bank statement scanning copy while monitoring whether the rectangular box coincides with a textbox on the bank statement scanning copy or not; and taking content information corresponding to the textbox as a required bank statement record when the textbox coincides with the rectangular box. According to the method, all statement records in the whole file are subjected to structured extraction and a table is exported by using an automatic template sliding matching technology.

Description

technical field [0001] The invention belongs to the field of natural language processing, and in particular relates to a method for extracting paper bank flow information by using sliding template technology. Background technique [0002] In the financial industry, bank statement, as an important financial document, can objectively reflect the cash flow of individuals or enterprises, and has become an important document in economic activities such as credit, investment, and bond issuance. In addition to corporate accounts that can get electronic form version of the flow, other scenarios such as personal flow are generally paper flow, and it is impossible to directly obtain structured information for processing and analysis. In addition, due to the large number of banks and different software and hardware equipment, resulting in various flow formats, most scenarios for extracting paper-based bank flow information are still in the manual entry stage. [0003] With the continu...

AI Technical Summary

Problems solved by technology

Due to the large number and types of flow records, the flow review takes a lot of manpower and energy, and often requires a large number of practitioners to carry out boring electronic entry work

[0005] Due to the large number of flow record files, which are affected by various factors such as fatigue, attention, and environment, it is easy to cause problems such as missing information and wrong records during the flow review and entry process, which in turn leads to inaccurate and incomplete follow-up reviews. , which poses a significant risk to the project

Images