A dual-recording operation mode recognition method based on OCR automatic recognition

Abstract

The application provides a double-recording operation mode recognition method based on OCR automatic recognition, and the recognition method comprises the following steps: acquiring transaction data; performing data verification, if the data verification is passed, downloading an image and performing OCR character recognition; otherwise, performing image data verification; determining whether to perform an image process, if yes, performing image segmentation; otherwise, performing image data verification; determining whether to perform a data entry process, if yes, performing a first entry process and a second entry process respectively; otherwise, performing image data verification; judging whether the results of the first entry process and the second entry process are consistent, if yes, performing image data verification, and if not, performing image data verification. Through the joint of the slice service and the OCR image recognition technology, the rapid entry of image information is realized, and the speed is fast and the accuracy is high.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of automatic identification, in particular to a double-recording operation mode identification method based on OCR automatic identification. BACKGROUND

[0002] When a bank processes bill-type business such as nationwide check image business, payment business, and city business through a centralized operation system, it often adopts an image slicing processing and information fragmentation recording mode to process the flow operation mode. In the information recording process, a two-recording-one-check mode is generally adopted, that is, two fragment recording personnel are required to record each fragment, and if the two recordings are consistent, it is passed, and if the two recordings are inconsistent, an arbitrator is required to arbitrate. The arbitrator completes the fragment recording link by selecting one of the recordings or recording himself.

[0003] At present, fragment recording is still simply manually recorded through a front-end web interface, and has not been connected with OCR image recognition technology. Therefore, at least two employees are required to complete the fragment recording link, which is low in business handling efficiency, high in overall operation cost, poor in customer experience, and low in comprehensive benefits of hall marketing.

[0004] The traditional image information recording process is to compare the fragment images after being cut by artificial twice recording, and if the comparison results are different, a third person is required to arbitrate to determine the final value of the fragment. SUMMARY

[0005] In view of the above problems, the present application is proposed to provide a double-recording operation mode identification method based on OCR automatic identification to overcome the above problems or at least partially solve the above problems.

[0006] According to one aspect of the present application, a double-recording operation mode identification method based on OCR automatic identification is provided, and the identification method comprises:

[0007] acquiring transaction data;

[0008] performing data verification, if passed, downloading images, and performing OCR text recognition; otherwise, performing image data verification;

[0009] determining whether to perform an image process, if yes, performing image slicing; otherwise, performing image data verification;

[0010] determining whether to perform a data recording process, if yes, performing a first recording process and a second recording process respectively; otherwise, performing image data verification;

[0011] judging whether the results of the first recording process and the second recording process are consistent, if yes, performing image data verification, if not, performing data arbitration.

[0012] Optionally, the image data verification further comprises:

[0013] transaction data verification;

[0014] data return.

[0015] Optionally, the OCR text recognition specifically comprises:

[0016] According to the identification type and element field of the centralized job slice service, a model is established and model learning is performed to improve the identification accuracy;

[0017] According to the writing method, format and length requirement of the element field of the certificate, the identification rule is limited to further improve the identification accuracy.

[0018] Optionally, the OCR text recognition further comprises:

[0019] The slice service and the OCR customized interaction interface are connected through the ESB platform and the file transmission platform, and the certificate type to be identified and the corresponding certificate image file are transmitted to the OCR;

[0020] The OCR returns the identification content to the slice service according to the model of each certificate type;

[0021] The slice service parses the OCR identification result according to the certificate field mapping table.

[0022] Optionally, the slice service sends an interface request to the OCR after downloading the image, and stores the returned result in the entry element table in advance.

[0023] Optionally, the slice service further comprises: the slice service retrieves the corresponding OCR identification result in the entry element table for element comparison after the manual first recording is completed.

[0024] The application provides a double-recording job mode identification method based on OCR automatic identification, which comprises the following steps: acquiring transaction data; performing data verification; if the data verification is passed, downloading the image and performing OCR text recognition; otherwise, performing image data verification; determining whether to perform image flow; if yes, performing image segmentation; otherwise, performing image data verification; determining whether to perform data entry flow; if yes, performing the first entry flow and the second entry flow respectively; otherwise, performing image data verification; judging whether the results of the first entry flow and the second entry flow are consistent; if yes, performing image data verification; if not, performing data arbitration. Through the connection of the slice service and the OCR image recognition technology, the image information is quickly entered, the speed is fast, the accuracy is high, the business handling efficiency is greatly improved, the overall operation cost is reduced, the customer waiting time is reduced, the customer experience is improved, and the comprehensive benefits of hall marketing are improved.

[0025] The above description is only a summary of the technical solutions of the present application. In order to make the technical means of the present application more clearly understood and implemented according to the content of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application are described below. BRIEF DESCRIPTION OF DRAWINGS

[0026] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0027] Figure 1 A flowchart of a double-recording operation mode recognition method based on OCR automatic recognition provided by the embodiment of the present application. DETAILED DESCRIPTION

[0028] The exemplary embodiments of the present disclosure will be described in detail with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be accurately conveyed to those skilled in the art.

[0029] The terms "include" and "have" and any variations thereof in the specification, claims and drawings of the present application are intended to cover non-exclusive inclusion, for example, including a series of steps or units.

[0030] The technical solutions of the present application will be further described in detail below in combination with the drawings and embodiments.

[0031] As shown in Figure 1 A double-recording operation mode recognition method based on OCR automatic recognition includes: acquiring transaction data; performing data verification, if passed, downloading images and performing OCR text recognition; otherwise, performing image data verification; determining whether to perform image flow, if yes, performing image segmentation; otherwise, performing image data verification; determining whether to perform data entry flow, if yes, performing first and second entry flows respectively; otherwise, performing image data verification; determining whether the results of the first and second entry flows are consistent, if yes, performing image data verification, if not, performing data arbitration.

[0032] OCR (Optical Character Recognition) is a text recognition technology that converts text in scanned images into text information. Integrating OCR text recognition technology into image fragment input allows for the rapid and accurate acquisition of image information, greatly reducing errors caused by manual input and ensuring the accuracy of the information.

[0033] The integration solution between centralized slicing services and OCR is achieved through the following steps:

[0034] Based on the types of vouchers and element fields that the centralized operation slicing service needs to identify, OCR pre-builds models and performs model learning to improve recognition accuracy. OCR can restrict recognition rules according to the writing style, format, length and other requirements of voucher element fields, such as handwritten, machine-printed, amount, date, 8-digit voucher number, etc., to further improve recognition accuracy.

[0035] The slicing service interacts with the OCR customized interface through the ESB platform and file transfer platform. The type of voucher to be recognized and the corresponding voucher image file are transmitted to the OCR. The OCR returns the recognized content to the slicing service according to the model of each voucher type. The slicing service parses the OCR recognition result according to the voucher field mapping table.

[0036] After downloading the image, the tiling service sends an interface request to the OCR and pre-stores the returned results in the input element table.

[0037] After manual data entry is completed, the slice service retrieves the corresponding OCR recognition results from the entered element table for element comparison.

[0038] Beneficial effects: By integrating the slicing service with OCR image recognition technology, rapid image information input is achieved. This is fast, accurate, and significantly improves business processing efficiency, reduces overall operating costs, reduces customer waiting time, improves customer experience, and enhances the overall effectiveness of in-store marketing.

[0039] The above specific embodiments further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for recognizing dual-recording operation patterns based on OCR automatic recognition, characterized in that, The identification method includes: Obtain transaction data; Perform data verification. If it passes, download the image and perform OCR text recognition; otherwise, perform image data verification. The OCR text recognition specifically includes: Based on the voucher types and element fields that the centralized operation slicing service needs to identify, a model is pre-built and model learning is performed to improve the recognition accuracy. Based on the writing style, format, and length requirements of the voucher element fields, the recognition rules are restricted to further improve the recognition accuracy. The slicing service and OCR customization interaction interface transmit the voucher type to be recognized and the corresponding voucher image file to the OCR through the ESB platform and file transfer platform; OCR returns the identified content to the slicing service based on the model for each credential type; The slicing service parses the OCR recognition results based on the credential field mapping table; Determine whether an image processing workflow is required. If so, perform image segmentation; otherwise, perform image data verification. Determine whether a data entry process is required. If so, proceed with the first and second data entry processes respectively; otherwise, perform image data verification. Determine whether the results of the first data entry process and the second data entry process are consistent. If they are consistent, perform image data verification. If they are inconsistent, perform data arbitration.

2. The dual-recording operation mode recognition method based on OCR automatic recognition according to claim 1, characterized in that, The image data verification process also includes: Transaction data verification; Data feedback.

3. The dual-recording operation mode recognition method based on OCR automatic recognition according to claim 1, characterized in that, After downloading the image, the slicing service sends an interface request to the OCR and stores the returned result in the input element table in advance.

4. The dual-recording operation mode recognition method based on OCR automatic recognition according to claim 1, characterized in that, The slicing service also includes: after the manual recording is completed, the slicing service retrieves the corresponding OCR recognition results in the entered element table for element comparison.

Images