WO2019161615A1

WO2019161615A1 - Bill entry method, system, optical character recognition server and storage medium

Info

Publication number: WO2019161615A1
Application number: PCT/CN2018/083768
Authority: WO
Inventors: 李响
Original assignee: 平安科技（深圳）有限公司
Priority date: 2018-02-23
Filing date: 2018-04-19
Publication date: 2019-08-29
Also published as: CN108346106A

Abstract

Disclosed is a bill entry method, said bill entry method comprising the following steps: when a bill identification request sent by an entry device is received, acquiring a bill image according to the bill identification request; determining an identification area of the bill image and identifying text information in the identification area; and feeding back the text information to the entry device for said entry device to enter the text information. Further disclosed are a bill entry system, an optical character recognition server, and a storage medium. Due to the optical character recognition server, the present application only needs to identify the text information of the identification area, and does not need to globally identify the bill, which reduces the identification amount of the bill image and reduces the length of time spent identifying the text information, thereby improving the entry efficiency of the bill.

Description

Bill entry method, system, optical character recognition server and storage medium

This application claims the priority of the Chinese patent application filed on February 23, 2018, the Chinese Patent Office, application number 201810155306.2, and the invention name is "bill entry method, system, optical character recognition server and storage medium". The citation is incorporated in the application.

Technical field

The present application relates to the field of communications technologies, and in particular, to a bill entry method, system, optical character recognition server, and storage medium.

Background technique

As people's living standards improve, more and more people have the ability to purchase insurance to reduce the loss of accidents.

After the user who purchases the insurance gets the claim statement, he needs to go to the claim office to settle the claim amount, and the customer service needs to enter the claim statement into the system at the same time as the claim settlement, to calculate the claim amount.

In the prior art, the entry of the claim statement is manually entered by the customer service, and the error rate of the entry method is higher, and the entry time is longer, so that the entry efficiency of the claim statement is low.

Summary of the invention

The main purpose of the present application is to provide a bill entry method, system, optical character recognition server, and storage medium, which are intended to solve the problem of low entry efficiency of bills.

To achieve the above objective, the present application provides a bill entry method, and the bill entry method includes the following steps:

Upon receiving the bill identification request sent by the entry device, obtaining a bill image according to the bill identification request;

Determining an identification area of the bill image and identifying text information in the identification area;

The text information is fed back to the entry device for the entry device to enter the text information.

Preferably, the step of determining the identification area of the bill image includes:

Determining the type of the billing image;

An identification area of the billing image is determined according to the type.

Preferably, the step of determining the type of the billing image comprises:

Determining a preset identification area of the bill image;

Identifying the preset identification area to obtain identification information of the bill;

The type of the billing image is determined based on the identification information.

Preferably, the step of identifying the text information in the identification area comprises:

Determining that the identification area is at a first location of the billing image;

Determining an identification object of the identification area according to the type and the first position, wherein the identification object includes at least one of a character, a number, a letter, and a formula;

Identifying text information in the identified object.

Preferably, after the step of identifying the text information in the identification area, the method further includes:

Obtaining a document template corresponding to the billing image according to the type;

Determining, according to the first location, the text information in a second location of the document template;

Adding the text information to the second location in the document template to obtain a billing document;

The billing document is fed back to the entry device to cause the entry device to enter the text information according to the bill document.

Preferably, the step of determining, according to the first location, the text information in the second location of the document template comprises:

Obtaining a position mapping relationship between the bill image and the document template;

The second location is determined according to the first location and the location mapping relationship.

Preferably, the step of obtaining a bill image according to the bill identification request comprises:

Obtaining a pending billing image according to the bill identification request;

Obtaining a preset grayscale threshold, and binarizing the to-be-processed billing image according to the preset grayscale threshold to obtain the billing image.

To achieve the above object, the present application also provides an optical character recognition server including a processor, a memory, and a bill entry program stored on the memory and operable on the processor, The step of the bill entry method as described above is implemented when the bill entry program is executed by the processor.

To achieve the above object, the present application also provides a storage medium storing a bill entry program, the bill entry program being executed by the processor to implement the steps of the bill entry method as described above.

In order to achieve the above object, the present application further provides a bill entry system, the bill entry system comprising an optical character recognition server and at least one entry device, wherein

The optical character recognition server is configured to: when receiving the bill identification request sent by the input device, acquire a bill image according to the bill identification request, determine an identification area of the bill image, and identify text information in the identification area, Returning the text information to the input device for the input device to input the text information;

The input device is configured to receive text information sent by the optical character recognition server, determine an input template corresponding to the text information, and add the text information to the input template, where the input device receives the bill When the request is entered, the bill is scanned to obtain a bill image, and a bill identification request containing the bill graph is sent to the optical character recognition server for the optical character recognition server to feed back the text information in the bill image to the Record the device.

The present application provides a bill entry method, system, optical character recognition server, and storage medium. When receiving a bill identification request sent by an entry device, the bill image is acquired according to the bill identification request, and the identification area of the bill image is determined, and the identification area is identified. In the text information, the text information is fed back to the input device for inputting the text information of the billing device; since the optical character recognition server only needs to recognize the text information of the identification area, the bill is not required to be globally recognized, and the statement is reduced. The recognition amount of the image reduces the length of time for identifying the text information, thereby improving the efficiency of bill entry.

DRAWINGS

1 is a structural diagram of a system involved in an embodiment of the present application;

2 is a schematic structural diagram of hardware of an optical character recognition server according to an embodiment of the present application;

3 is a schematic flow chart of a first embodiment of a bill entry method of the present application;

4 is a schematic diagram of a refinement process of step S10 in FIG. 3;

FIG. 5 is a schematic diagram of a refinement process for determining an identification area of the bill image in step S20 of FIG. 3;

6 is a schematic flow chart of a second embodiment of a bill entry method of the present application;

FIG. 7 is a schematic flowchart diagram of a third embodiment of a bill entry method according to the present application. .

The implementation, functional features and advantages of the present application will be further described with reference to the accompanying drawings.

Detailed ways

It is understood that the specific embodiments described herein are merely illustrative of the application and are not intended to be limiting.

The main solution of the embodiment of the present application is: when receiving the bill identification request sent by the input device, acquiring the bill image according to the bill identification request; determining the identification area of the bill image, and identifying the text in the recognition area Information; feeding back the text information to the entry device for the entry device to enter the text information.

Because in the prior art, the entry of the claim statement is manually entered by the customer service, such an entry method has a higher error rate and a longer entry time, which makes the entry of the claim statement less efficient.

The present application provides a solution, because the optical character recognition server only needs to identify the text information of the identification area, and does not need to globally identify the bill, thereby reducing the recognition amount of the billing image and reducing the length of time for identifying the text information, thereby improving The efficiency of entry of bills.

As shown in FIG. 1 , the system architecture involved in the method of this embodiment includes an optical character recognition server and an input device, where

As an implementation, the hardware structure of the above optical character recognition server may be as shown in FIG. 2 .

Referring to FIG. 2, the optical character recognition server may include a processor 1001, such as a CPU, a memory 1002, and a communication bus 1003. The communication bus 1003 is configured to implement connection communication between components in the optical character recognition server. The memory 1002 may be a high speed RAM memory or a stable memory (non-volatile) Memory), such as disk storage. The memory 1002 can also optionally be a storage device independent of the aforementioned processor 1001. As shown in FIG. 2, a bill entry program may be included in the memory 1002 as a computer storage medium.

The processor 1001 in the hardware involved in the optical character recognition server shown in FIG. 2 can be used to call the bill entry program stored in the memory 1002 to execute the bill entry method described in the embodiment.

Based on the above hardware architecture, an embodiment of the bill entry method of the present application is proposed.

Referring to FIG. 3, FIG. 3 is a first embodiment of a bill entry method according to the present application. The bill entry method includes the following steps:

Step S10, when receiving the bill identification request sent by the input device, acquiring a bill image according to the bill identification request;

In the present application, the execution subject is an optical character recognition server (OCR server). The OCR server is connected to a plurality of input devices, and the input device is provided with a scanning module capable of scanning the claim statement to obtain a bill image, and then the input device sends a bill identification request containing the bill image to the OCR server. It should be noted that the OCR server is When the billing identification request is received, the image of the bill to be processed is obtained, and the image of the bill to be processed needs to be binarized. Specifically, please refer to FIG. 4, that is, step S10 includes:

Step S11, acquiring a to-be-processed billing image according to the bill identification request;

The input device stores the image of the claim statement uploaded by the user. When the user needs to perform the settlement of the claim, the corresponding claim statement image may be clicked on the input device, and the input device generates the claim when the user clicks on the claim statement image operation. The bill identification request of the bill image, and the bill identification request is sent to the OCR server, the OCR server parses the bill identification request, and obtains the claim bill image, and the claim bill image is the pending bill image.

Step S12, acquiring a preset grayscale threshold, and binarizing the to-be-processed billing image according to the preset grayscale threshold to obtain the billing image;

After obtaining the to-be-processed billing image carried in the bill identification request, the OCR server obtains a preset grayscale threshold (the preset grayscale threshold may be any suitable value), and binarizes the pending bill according to the preset grayscale threshold. An image to obtain a binary image (bill image) for the OCR server to recognize the bill image.

It should be noted that when multiple input devices are connected to the OCR server, the input device may perform binarization processing on the scanned bill image to obtain a binarized bill image, and then send a binned bill image to the OCR server. Bill identification request; because the binarized bill image has a much smaller capacity than the scanned bill image, thereby reducing the amount of input to the OCR server, enabling the OCR server to maintain sufficient bandwidth to ensure compatibility with each entry device The data exchange rate, which guarantees that the bill entry time is not affected by the network conditions.

Step S20, determining an identification area of the bill image, and identifying text information in the identification area;

The claim statement consists of multiple areas and blank parts. The area contains text information, that is, the OCR server only needs to recognize the text information of the area, and does not need to identify the blank area. Based on this, the claim bill can set a plurality of fixed location areas, which are used to fill in text information, and the OCR server only needs to identify the text information in these areas. Specifically, please refer to FIG. 5, that is, step S20 includes:

Step S21, determining a type of the billing image;

Each type of claim statement includes corresponding identification information, and the identification information may be a number consisting of characters, letters, and numbers. The identification information is set in a fixed area of the claim statement (preset identification area), and the OCR server only needs to be in the fixed area. The identification information is identified to obtain the number of the claim statement, and the type of the claim statement is determined according to the number.

Step S22, determining an identification area of the bill image according to the type;

Different types of claims bills have different locations and areas for filling out texts, that is, each type of claim bill has a fixed text field, and the OCR server stores various types of claim bills. Fill in the text area (identification area); when receiving the bill identification request, only need to determine the type of the bill image in the bill identification request, and then determine the identification area of the bill image according to the type, thereby the text information in the identification area Identify.

Step S30, feeding back the text information to the input device, so that the input device inputs the text information;

After the OCR server recognizes the text information of the identification area, the text information is fed back to the input device, and the input device records the text information into the corresponding entry template to complete the entry of the claim statement. It should be noted that the entry template can be set according to the claim statement, that is, the entry interface of the entry template is consistent with the text filling face of the claim statement, and the OCR server can carry the text information to the input device while carrying the identification area corresponding to the text information. The identification is such that the input device records the text information into the entry position corresponding to the entry template according to the identification area identifier.

In the technical solution provided by the embodiment, when receiving the bill identification request sent by the input device, the OCR server obtains the bill image according to the bill identification request, determines the identification area of the bill image, and recognizes the text information in the identification area, and then The text information is fed back to the input device for inputting the text information of the billing device; the OCR server uses the OCR server to identify the text information of the claim bill to automatically enter the claim bill, thereby avoiding the problem of high error rate of manual entry of the claim bill, and The labor cost is reduced, and the efficiency of billing is improved. In addition, the OCR server only needs to identify the text information of the identification area, and does not need to globally identify the bill, thereby reducing the recognition amount of the bill image, thereby reducing the recognition text information. The length of time has further improved the efficiency of bill entry.

Referring to FIG. 6, FIG. 6 is a second embodiment of the bill entry method of the present application. Based on the foregoing embodiment, after the step S21, the step S20 further includes:

Step S23, determining that the identification area is at a first position of the billing image;

The text information in the claim statement is not limited to words, but also numbers, letters and formulas, and the input device only needs to enter words, numbers, letters or formulas; and the OCR server only needs to recognize numbers, letters or formulas, and does not need all Identification. Based on this, the OCR server stores an identification object for each identification area of each type of bill (the identification object includes at least one of characters, numbers, letters, and/or formulas), that is, the OCR server sets the billing type, the identification area of the bill. The location and the mapping relationship of the identified objects. After the OCR server determines the type of the bill, it is determined that the identified area in the bill is at the position of the bill (first position).

Step S24, determining an identification object of the identification area according to the type and the first location, where the identification object includes at least one of a character, a number, a letter, and a formula;

Step S25, identifying text information in the identification object;

The OCR server determines the identification object of the identification area according to the first location and the bill type, and then identifies the identification object to obtain corresponding text information. For example, the claims bill that needs to be identified contains three areas: A, B, and C. Area A is used to fill out the claims and the amount of claims corresponding to the claims, that is, area A consists of words and numbers; area B is used to calculate the amount of claims. Principle and calculation formula, B area is composed of words, numbers, letters and formulas. Area C is used to fill in the matters needing attention when the claim is settled, that is, the C area only contains text; and the A' corresponding to the A area in the entry template of the input device The area needs to fill in the text and numbers. The B' area corresponding to the B area only needs to fill in the formula. The C' area corresponding to the C area only needs to fill in the number; the OCR server determines the position of the A, B, and C areas, and then according to the position (the first A position) identifies the character and number of the A area, the formula for identifying the B area, and the number identifying the C area.

It should be noted that if the identification object is only one of the text information (such as a number), the number needs to be separated; in general, there is no character interval between the respective numbers of the complete value, and when the character interval occurs, Separate numbers (such as commas instead of character spacing) to form two values, and so on, text, letters, and formulas can be separated in the manner described above.

In the technical solution provided by the embodiment, after determining the identification area of the bill image, the OCR server further determines the location of the identification area, thereby determining the identification object of the identification area according to the location; because the identification object is text, numbers, letters and At least one of the formulas further reduces the recognition amount of the OCR server and improves the efficiency of the entry of the claims bill.

Referring to FIG. 7, FIG. 7 is a third embodiment of the present application. After the step S20, the method further includes:

Step S40: Obtain a document template corresponding to the billing image according to the type;

In order to facilitate entry of the text information by the entry device, the OCR server can typeset the text information of each identification area of the bill image. The OCR server sets a corresponding document template for each type of bill. After the OCR server determines the type of the bill, the corresponding document template can be determined according to the type of the bill.

Step S50, determining, according to the first location, that the text information is in a second location of the document template;

In order to facilitate adding the text information of the identification area to the corresponding position in the document template, the OCR server sets the mapping relationship between the location of the identification area of the bill and the text addition position of the document template, when the OCR server performs the identification of the text information of the identification area, according to The position (first position) of the recognition area and the mapping relationship determine the position (second position) of the text information in the document template.

Step S60, adding the text information to the second location in the document template to obtain a billing document;

After determining the location of the textual information in the document template, the textual information is added to the second location in the document template.

Step S70, feeding back the billing document to the input device, so that the input device inputs the text information according to the billing document;

After the OCR server adds all the text information to the document template, the bill document is obtained. At this time, the bill document is sent to the input device, so that the input device reads the text information from the bill document, and then enters the read text information. Enter the template.

Further, the document template and the input template of the input device also have a position mapping relationship, so that the input device records the text information in the document template into the template through the position mapping relationship, thereby completing the entry of the claim statement. Of course, the document template of the OCR server can be directly used as the input template of the input device, that is, after the text information is entered in the document template, the OCR server sends the bill document (the document template after the text information is entered) to the input device, and the input device saves the file template. The billing document is fine.

In the technical solution provided by the embodiment, after identifying the text information of the identification area, the OCR server determines the second position of the text information on the document template according to the identification area in the first position of the billing image, so that according to the second position, The text information is added to the document template for the input device to input the text information according to the document template; the OCR server formats the recognized text information, thereby facilitating the entry system to input the typed text information, thereby improving the entry of the claim statement. effectiveness.

The present application also provides an optical character recognition server including a processor, a memory, and a bill entry program stored on the memory and operable on the processor, the bill entry program being The steps of the bill entry method described in the above embodiment are implemented when the processor is executed.

To achieve the above object, the present application also provides a storage medium storing a bill entry program, the bill entry program being executed by the processor to implement the steps of the bill entry method as described in the above embodiments.

The serial numbers of the embodiments of the present application are merely for the description, and do not represent the advantages and disadvantages of the embodiments.

It is to be understood that the term "comprises", "comprising", or any other variants thereof, is intended to encompass a non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes those elements. It also includes other elements that are not explicitly listed, or elements that are inherent to such a process, method, article, or device. An element that is defined by the phrase "comprising a ..." does not exclude the presence of additional equivalent elements in the process, method, item, or device that comprises the element.

Through the description of the above embodiments, those skilled in the art can clearly understand that the foregoing embodiment method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be through hardware, but in many cases, the former is better. Implementation. Based on such understanding, the technical solution of the present application, which is essential or contributes to the prior art, may be embodied in the form of a software product stored in a storage medium (such as ROM/RAM as described above). , a disk, an optical disk, including a number of instructions for causing a terminal device (which may be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to perform the methods described in the various embodiments of the present application.

The above is only a preferred embodiment of the present application, and is not intended to limit the scope of the patent application, and the equivalent structure or equivalent process transformations made by the specification and the drawings of the present application, or directly or indirectly applied to other related technical fields. The same is included in the scope of patent protection of this application.

Claims

A bill entry method, characterized in that the bill entry method comprises the following steps:

Upon receiving the bill identification request sent by the entry device, obtaining a bill image according to the bill identification request;

Determining an identification area of the bill image and identifying text information in the identification area;

The text information is fed back to the entry device for the entry device to enter the text information.
The bill entry method according to claim 1, wherein said determining said identification area of said bill image comprises:

Determining the type of the billing image;

An identification area of the billing image is determined according to the type.
The bill entry method according to claim 2, wherein said determining said type of billing image comprises:

Determining a preset identification area of the bill image;

Identifying the preset identification area to obtain identification information of the bill;

The type of the billing image is determined based on the identification information.
The bill entry method according to claim 2, wherein the step of identifying the text information in the identification area comprises:

Determining that the identification area is at a first location of the billing image;

Determining an identification object of the identification area according to the type and the first position, wherein the identification object includes at least one of a character, a number, a letter, and a formula;

Identifying text information in the identified object.
The bill entry method according to claim 4, wherein the step of identifying the text information in the identification area further comprises:

Obtaining a document template corresponding to the billing image according to the type;

Determining, according to the first location, the text information in a second location of the document template;

Adding the text information to the second location in the document template to obtain a billing document;

The billing document is fed back to the entry device to cause the entry device to enter the text information according to the bill document.
The bill entry method according to claim 5, wherein the step of determining the text information in the second position of the document template according to the first location comprises:

Obtaining a position mapping relationship between the bill image and the document template;

The second location is determined according to the first location and the location mapping relationship.
The bill entry method according to claim 1, wherein said step of obtaining a bill image according to said bill identification request comprises:

Obtaining a pending billing image according to the bill identification request;

Obtaining a preset grayscale threshold, and binarizing the to-be-processed billing image according to the preset grayscale threshold to obtain the billing image.
An optical character recognition server, comprising: a processor, a memory, and a bill entry program stored on the memory and operable on the processor, the bill entry program being The following steps are implemented when the processor executes:

Upon receiving the bill identification request sent by the entry device, obtaining a bill image according to the bill identification request;

Determining an identification area of the bill image and identifying text information in the identification area;

The text information is fed back to the entry device for the entry device to enter the text information.
The optical character recognition server according to claim 8, wherein said step of determining an identification area of said bill image comprises:

Determining the type of the billing image;

An identification area of the billing image is determined according to the type.
The optical character recognition server according to claim 9, wherein said determining said type of billing image comprises:

Determining a preset identification area of the bill image;

Identifying the preset identification area to obtain identification information of the bill;

The type of the billing image is determined based on the identification information.
The optical character recognition server according to claim 9, wherein said step of identifying text information in said identification area comprises:

Determining that the identification area is at a first location of the billing image;

Determining an identification object of the identification area according to the type and the first position, wherein the identification object includes at least one of a character, a number, a letter, and a formula;

Identifying text information in the identified object.
The optical character recognition server according to claim 11, wherein said step of recognizing said text information in said identification area is performed by said processor when said bill entry program is executed by said processor:

Obtaining a document template corresponding to the billing image according to the type;

Determining, according to the first location, the text information in a second location of the document template;

Adding the text information to the second location in the document template to obtain a billing document;

The billing document is fed back to the entry device to cause the entry device to enter the text information according to the bill document.
The optical character recognition server according to claim 12, wherein the step of determining the text information in the second position of the document template according to the first location comprises:

Obtaining a position mapping relationship between the bill image and the document template;

The second location is determined according to the first location and the location mapping relationship.
The optical character recognition server according to claim 8, wherein said step of obtaining a bill image according to said bill identification request comprises:

Obtaining a pending billing image according to the bill identification request;

Obtaining a preset grayscale threshold, and binarizing the to-be-processed billing image according to the preset grayscale threshold to obtain the billing image.
A storage medium, characterized in that the storage medium stores a bill entry program, and the bill entry program is executed by the processor to implement the following steps:

Upon receiving the bill identification request sent by the entry device, obtaining a bill image according to the bill identification request;

Determining an identification area of the bill image and identifying text information in the identification area;

The text information is fed back to the entry device for the entry device to enter the text information.
The storage medium according to claim 15, wherein said determining said identification area of said bill image comprises: determining a type of said bill image;

An identification area of the billing image is determined according to the type.
The storage medium of claim 16, wherein the step of determining the type of the billing image comprises:

Determining a preset identification area of the bill image;

Identifying the preset identification area to obtain identification information of the bill;

The type of the billing image is determined based on the identification information.
The storage medium according to claim 16, wherein said step of identifying text information in said identification area comprises:

Determining that the identification area is at a first location of the billing image;

Determining an identification object of the identification area according to the type and the first position, wherein the identification object includes at least one of a character, a number, a letter, and a formula;

Identifying text information in the identified object.
The storage medium according to claim 18, wherein said step of recognizing said text information in said identification area is performed by said processor when said bill entry program is executed by said processor:

Obtaining a document template corresponding to the billing image according to the type;

Determining, according to the first location, the text information in a second location of the document template;

Adding the text information to the second location in the document template to obtain a billing document;

The billing document is fed back to the entry device to cause the entry device to enter the text information according to the bill document.
A bill entry system, characterized in that the bill entry system comprises an optical character recognition server and at least one entry device, wherein

The optical character recognition server is configured to: when receiving the bill identification request sent by the input device, acquire a bill image according to the bill identification request, determine an identification area of the bill image, and identify text information in the identification area, Returning the text information to the input device for the input device to input the text information;

The input device is configured to receive text information sent by the optical character recognition server, determine an input template corresponding to the text information, and add the text information to the input template, where the input device receives the bill When the request is entered, the bill is scanned to obtain a bill image, and a bill identification request containing the bill graph is sent to the optical character recognition server for the optical character recognition server to feed back the text information in the bill image to the Record the device.