CN115731377A - Bidder information verification system, method and device based on picture identification - Google Patents

Abstract

The invention discloses a bidder information verification system, method and device based on picture recognition, including: acquiring bidder's identity information picture; performing preprocessing and identification to obtain bidder's unified social credit code, legal representative name and Its citizen ID number, the name of the person in charge of the project and its citizen ID number; use a third-party credit agency to verify whether there is a record of bad faith and bribery; when the verification fails, extract the identification result and send It is stored in the manual verification data table; when the verification is passed, the legal person certificate and business license of the bidder are stored in the verification data table; the present invention feedbacks whether the bidder exists by identifying the legal person certificate or business license information provided by the bidder. Bribery crime records, bad integrity records, etc.; solve the technical problems that the bidding system verification process cannot be verified in real time, and it takes a lot of manpower and time to verify.

Description

A bidder information verification system, method and device based on image recognition

technical field

The invention belongs to the technical field of information processing, and in particular relates to a system, method and device for verifying bidder information based on image recognition.

Background technique

In the process of verifying the identity of the bidding company, it is necessary to determine whether the bidding company has a record of bribery crimes, bad integrity and other information, which specifically includes: 1. Searching for bidders, legal representatives and projects through the "China Judgment Documents Network" Whether the person in charge (such as the person in charge of the project required to be provided in the bidding documents) has a record of bribery within the effective date specified in the bidding documents, etc. 2. To inquire about bidders' poor integrity records, and divide them into institutions and enterprises (non-institutions) according to the nature of bidders. For public institutions, it is necessary to check the basic information of bidders and whether there is any administrative penalty within the effective date stipulated in the bidding documents from the "Institution Online"; for enterprises (non-public institutions), it is necessary to query the bids from the "National Enterprise Credit Information Publicity System" Whether the person is "included in the list of seriously illegal and untrustworthy enterprises (blacklist)" and "included in the list of abnormal business operations" within the effective date specified in the bidding documents. 3. Use "Credit China" to check whether the bidder has a "blacklist" record within the effective date specified in the bidding documents.

In the bid evaluation process, it is first necessary to judge whether it is a public institution or an enterprise (non-public institution), and check whether the bidder has a bad record of integrity in the "Institution Online" and "National Enterprise Credit Information Publicity System" respectively. Name or unified social credit code to check whether there is a criminal record of bribery, associate the bidder’s name or unified social credit code with the name of the legal representative, and the citizen’s identity number to check whether the legal representative has a criminal record of bribery, and use the bidder’s name or unified social credit The code is associated with the project leader’s name and citizen ID number to check whether the project leader has a criminal record of bribery. The above verification process cannot be verified in real time, and it takes a lot of manpower and time to verify.

Contents of the invention

The technical problem to be solved by the present invention is to provide a bidder information verification system, method and device based on image recognition, so as to solve the technical problems that the verification process of the bidding system cannot be verified in real time, and a large amount of manpower and time are required for verification. .

Technical scheme of the present invention:

A method for verifying bidder information based on image recognition, the method comprising:

Obtain the identity information picture of the bidder; the identity information picture includes the picture of the legal person certificate, the business license picture, the legal representative ID card picture and the project leader's ID card picture;

Preprocess the bidder’s identity information image to obtain the corrected image of the bidder’s identity information; preprocessing includes tilt correction, grayscale, USM sharpening and binarization;

Identify the corrected image of the bidder's identity information, and obtain the bidder's unified social credit code, legal representative name and citizen ID number, project leader's name and citizen ID number;

Obtain the effective date stipulated in the bidding documents, and when the current date is within the effective date, use a third-party credit investigation agency to verify the bidder to determine whether the bidder has a record of bad faith and bribery; the third-party credit investigation agency Provide China Judgment Documents Network, Public Institution Online or National Enterprise Credit Information Publicity System;

When the verification fails, extract the bidder's unified social credit code, legal representative's name and citizen ID number, project leader's name and citizen ID number, and store them in the manual verification data table for manual verification;

When the verification is passed, the legal person certificate and business license of the bidder shall be stored in the verification data sheet.

Before the step of obtaining the bidder's identification information picture, also include:

Obtain images of legal person certificates or business licenses of public institutions, enterprises, institutions, and enterprises from bidding documents and store them in the bidder qualification certificate image library, obtain legal representative ID card images from bidding documents and store them in the legal representative ID card Image library, obtain the ID card picture of the project leader from the bidding documents and store it in the ID card image library of the project leader.

The step of preprocessing the bidder's identity information picture to obtain the bidder's identity information corrected picture includes:

When the bidder's identity information picture has an inclination, correct the inclination of the bidder's identity information picture to obtain a first processed picture;

When the identity information picture of the bidder has no inclination or after obtaining the first processed picture, the RGB value of each pixel in the identity information picture or the first processed picture is mapped to a gray value of 0-255 to obtain gray processing pictures;

Record the segmentation threshold of the foreground and background of the gray scale processing picture as T, calculate the variance value, and make the segmentation threshold T traverse 0-255, calculate the variance value each time, and extract the T value that maximizes the variance value as target threshold;

Using the USM sharpening algorithm to sharpen the grayscale processed picture by using the target threshold to obtain the sharpened processed picture;

Perform binarization on the sharpened image to obtain the corrected image of the bidder's identity information.

The segmentation threshold of the foreground and background of the grayscale processing picture is denoted as T, and in the step of calculating the variance value, the calculation formula is:

为属于前景的像素点数占整幅图片的比例，平均灰度为u₀；

为背景像素点数占整幅图片的比例，平均灰度为u₁。Among them, g is the variance value, M×N is the size of the picture, N ₀ is the number of pixels whose gray value of the pixel value in the picture is less than the threshold T, N ₁ is the number of pixels in the picture whose gray level is greater than the threshold T,

is the proportion of pixels belonging to the foreground to the whole picture, and the average gray level is u ₀ ;

is the proportion of the number of background pixels to the whole picture, and the average gray level is u ₁ .

The step of binarizing the sharpened picture to obtain the corrected picture of the bidder's identity information includes:

Scan the sharpened image horizontally, and add up the gray values of all points with light colors on both sides and dark colors in the middle of the sharpened image as sum ₁ , and record the accumulated points as p ₁ ;

Scanning the sharpened picture vertically, and summing up the gray values of all points whose upper and lower sides of the sharpened picture are light-colored and the middle is dark-colored are recorded as sum ₂ , and the accumulated points are recorded as p ₂ ;

Calculate the threshold value Y of binarization; wherein the calculation formula is: Y=(sum ₁ +sum ₂ )/(p ₁ +p ₂ );

Using the threshold Y to binarize the sharpened picture to obtain a corrected picture of the bidder's identity information with the watermark removed.

The step of identifying the corrected image of the bidder's identity information to obtain the unified social credit code of the bidder, the name of the legal representative and its citizen ID number, the name of the project leader and its citizen ID number, includes:

Use the character recognition network model to judge the area where the unified social credit code is located on the legal person certificate picture or business license picture of the institution, enterprise, institution, and enterprise, and define this area as a character area, and define other areas as text areas;

Use the character recognition network model to identify the ID card picture of the legal representative and the ID card picture of the project leader, judge the area where the citizen ID number is located and define this area as a character area, and define other areas as text areas;

Use character recognition for the character area of the legal person certificate picture or business license picture of the public institution, enterprise institution, enterprise, and determine the legal representative of the bidder and the unified social credit code information;

Use character recognition on the character area of the legal representative's ID card picture and the project leader's ID card picture to determine the citizenship number of the legal representative of the bidder and the project leader;

Use OCR image recognition technology to identify the text in the text area of the legal person certificate picture or business license picture of the institution, enterprise institution, enterprise, and determine the name, category and legal representative information of the bidder;

Use OCR image recognition technology to identify the text in the text area of the legal representative's ID card picture and the project leader's ID card picture, and determine the name information of the legal representative of the bidder and the project leader.

When the verification is passed, after the step of storing the bidder’s legal person certificate and business license in the verification data table, it also includes:

Automatically enter the legal person certificate or business license information of public institutions, enterprises, institutions, and enterprises that pass the verification in the data sheet into the bid evaluation system, and automatically enter the legal person certificate pictures or business licenses of institutions, enterprises, institutions, and enterprises that fail manual verification The information is stored in the exception database.

A bidder information verification system based on image recognition, including:

The obtaining module is used to obtain the identity information picture of the bidder; wherein, the identity information picture includes a picture of a legal person certificate, a picture of a business license, a picture of the ID card of the legal representative, and a picture of the ID card of the project leader;

The preprocessing module is used to preprocess the bidder's identity information picture to obtain the bidder's identity information corrected picture; wherein, the preprocessing includes tilt correction, grayscale, USM sharpening and binarization;

The identification module is used to identify the corrected picture of the bidder's identity information, and obtain the bidder's unified social credit code, the name of the legal representative and the citizen's ID card number, the name of the project leader and the citizen's ID card number;

The verification module is used to obtain the effective date stipulated in the bidding documents, and when the current date is within the effective date, use a third-party credit investigation agency to verify the bidder, and determine whether the bidder has a record of bad faith and bribery; Among them, the third-party credit reporting agency provides China Judgment Documents Network, Public Institution Online or the National Enterprise Credit Information Publicity System;

The first storage module is used to extract the bidder's unified social credit code, legal representative name and citizen ID number, project leader's name and citizen ID number when the verification fails, and store them in the manual verification data table China and Israel for manual verification;

The second storage module stores the bidder's legal person certificate and business license in the verification pass data table when the verification is passed.

A computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the steps of the method when executing the computer program.

A computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method are implemented.

Beneficial effects of the present invention:

The present invention identifies the legal person certificate or business license information of the bidder through the system, and uses data such as the China Judgment Document Network, the National Enterprise Credit Information Publicity System, the Institution Online, and Credit China provided by the third credit agency to automatically identify the bidders. Verification, feedback to the bid evaluation experts whether the bidders and their legal representatives and project leaders have bribery crime records within the valid date stipulated in the bidding documents, whether the bidders have bad integrity records, etc., saving the time of bid evaluation experts and preventing man-made Operation errors ensure fairness and impartiality in the process of verifying bidder information, and assist bid evaluation experts to conduct efficient bid evaluation.

It solves the technical problems that the verification process of the bidding system cannot be verified in real time, and a large amount of manpower and time are required for the verification.

Description of drawings

Fig. 1 is a schematic flow chart of a method according to an embodiment of the present invention.

Fig. 2 is a schematic diagram of the device structure of an embodiment of the present invention.

FIG. 3 is a schematic diagram of the internal structure of a computer device according to an embodiment of the present invention.

Detailed ways

As shown in Fig. 1, the present invention provides a kind of method of bidder information check based on picture recognition, first judge whether the bidder is a public institution or an enterprise (non-public institution), and inquire respectively in the public institution online and the national enterprise credit information publicity system Whether there is any record of bad faith within the effective date stipulated in its bidding documents. Inquiry of bribery records can be queried in three modes: bidder’s name or unified social credit code; bidder’s name or unified social credit code and legal representative’s name and its citizenship number, bidder’s name or unified social credit The code, the name of the person in charge of the project and his or her citizenship number respectively inquire whether there is a criminal record of bribery within the effective date stipulated in the bidding documents.

The method provided by the invention specifically includes:

Before the step S1, it also includes:

S01. Obtain the image of the legal person certificate or business license of the institution/enterprise from the bidding documents and store it in the bidder qualification certificate image library, obtain the ID card image of the legal representative from the bidding documents and store it in the ID card image library of the legal representative , obtain the ID card picture of the project leader from the bidding documents and store it in the ID card image library of the project leader.

S1. Obtain the identity information picture of the bidder; wherein, the identity information picture includes a legal person certificate picture, a business license picture, a legal representative ID card picture, and a project leader ID card picture;

S2. Perform preprocessing on the bidder's identity information image to obtain a corrected image of the bidder's identity information; wherein, the preprocessing includes tilt correction, grayscale, USM sharpening, and binarization; that is, Perform preprocessing on the acquired public institution legal person certificate or business license image, legal representative ID card image, and project leader ID card image. Preprocessing can also include image scaling, filtering and denoising.

Step S2 specifically includes:

S21. When the bidder's identity information picture has an inclination, correct the inclination of the bidder's identity information picture to obtain a first processed picture;

S22. When the bidder's identity information picture has no inclination, or after obtaining the first processed picture, map the RGB value of each pixel in the identity information picture or the first processed picture to 0-255 The grayscale value of the grayscale processing image is obtained; the grayscale processing is to map the RGB value of each pixel in the image to the grayscale value of 0-255, and the obtained image is the same as a black and white photo, which can reduce the subsequent steps The amount of calculation, and make the picture more recognizable.

S23. Denote the segmentation threshold of the foreground and background of the grayscale processed picture as T, calculate the variance value, and make the segmentation threshold T traverse 0-255, calculate the variance value each time, and extract the maximum variance value The T value of T is used as the target threshold; the text in the picture with the watermark can be identified. Since the watermark is translucent and covered on the text data, the edge of the text under the watermark becomes blurred and difficult to recognize, so the picture needs to be sharpened , which can highlight the text in the picture, which is conducive to the separation of the text in the picture from the watermark.

USM sharpening can not only enhance the edge of one side of the text, but also significantly weaken the other side of the edge of the text, the text will be highlighted, and the image near the text will be weakened, thereby reducing the area of the translucent watermark. And there will be white borders on the edges of the text. The threshold value in the USM sharpening algorithm needs to be set. If the threshold value is too large, the sharpening effect will not be produced; if the threshold value is set too small, the edges of text and watermark will be enhanced at the same time. Therefore, the setting of the threshold includes:

Denote the segmentation threshold of the foreground and background of the grayscale processing picture as T, calculate the variance value, and make the segmentation threshold T traverse 0-255, calculate the variance value every time, extract the T that makes the variance value the largest The value is used as the target threshold; the calculation formula for calculating the variance value is:

为属于前景的像素点数占整幅图片的比例，其平均灰度为u₀，

为背景像素点数占整幅图片的比例，其平均灰度为u₁。Among them, g is the variance value, M×N is the size of the picture, N ₀ is the number of pixels whose grayscale value is less than the threshold T in the picture, N ₁ is the number of pixels in the picture whose grayscale is greater than the threshold T,

is the proportion of pixels belonging to the foreground to the whole picture, and its average gray level is u ₀ ,

is the proportion of the number of background pixels to the whole picture, and its average gray level is u ₁ .

S24. Using the USM sharpening algorithm to sharpen the grayscale processed image by using the target threshold, to obtain a sharpened processed image;

S25. Perform binarization processing on the sharpened image to obtain a corrected image of the bidder's identity information; the image of the main text can be obtained through the binarization processing.

Step S25 specifically includes:

S251. Transversely scan the sharpened picture, and add up the gray values of all points with light colors on both sides and dark colors in the middle of the sharpened picture as sum ₁ , and record the accumulated points as p ₁ ;

S252. Scan the sharpened picture vertically, and add up the gray values of all points whose upper and lower sides of the sharpened picture are light-colored and the middle is dark-colored as sum ₂ , and the cumulative number of points is recorded as p ₂ ;

S253. Calculate the threshold value Y of binarization; wherein the calculation formula is: Y=(sum ₁ +sum ₂ )/(p ₁ +p ₂ ); that is, the gray value of the black point that is differentiated by the white edge is accumulated , and its sum is divided by the number of black points to obtain the average gray value of the black points.

S254. Binarize the sharpened picture by using the threshold Y to obtain a corrected picture of the bidder's identity information with the watermark removed.

S3. Identify the corrected image of the bidder's identity information, and obtain the unified social credit code of the bidder, the name of the legal representative and its citizen ID number, the name of the project leader and its citizen ID number;

Step S3 specifically includes:

S31, using the character recognition network model to determine the area where the unified social credit code is located on the legal person certificate picture or business license picture of the institution/enterprise and defining this area as a character area, and defining other areas as text areas;

S32. Using the character recognition network model to identify the ID card picture of the legal representative and the ID card picture of the project leader, determine the area where the citizen ID number is located and define this area as a character area, and define other areas as a text area;

S33. Use character recognition on the character area of the legal person certificate picture or business license picture of the institution/enterprise to determine the legal representative of the bidder and the unified social credit code information;

S34. Use character recognition on the character area of the ID card picture of the legal representative and the ID card picture of the project leader to determine the citizenship numbers of the legal representative of the bidder and the person in charge of the project;

S35. Use OCR image recognition technology to identify the text in the text area of the legal person certificate image or business license image of the institution/enterprise, and determine the name, category, legal representative and other information of the bidder;

S36. Use OCR image recognition technology to identify the text in the text area of the legal representative's ID card picture and the project leader's ID card picture, and determine the name information of the legal representative of the bidder and the project leader.

S4. Obtain the effective date stipulated in the bidding documents, and when the current date is within the effective date, use a third-party credit agency to verify the bidder, and determine whether the bidder has a record of bad faith and bribery ; Among them, the third-party credit reporting agency provides China Judgment Documents Network, Public Institution Online, and National Enterprise Credit Information Publicity System;

S5. When the verification fails, extract the unified social credit code of the bidder, the name of the legal representative and its citizen ID number, the name of the project leader and its citizen ID number, and store them in the manual verification data table China and Israel for manual verification;

S6. When the verification is passed, the legal person certificate and business license of the bidder are stored in the verification data table.

S7. Automatically input the legal person certificate or business license information of the public institution/enterprise that passed the verification in the data table into the bid evaluation system, and store the legal person certificate picture or business license information of the public institution/enterprise that failed the manual verification in the exception database. The information in the bidding documents shall first be judged whether there is a criminal record of bribery within the effective date stipulated in the bidding documents through the database data of the third-party credit investigation agency's China Judgment Documents Network, Public Institution Online, National Enterprise Credit Information Publicity System, Credit China, etc. Bad integrity records, if any, shall be stored in the database of invalid bids.

The present invention automatically verifies the bidder by identifying the bidder's public institution legal person certificate or business license information, using data such as the China Judgment Document Network, the National Enterprise Credit Information Publicity System, the public institution online, and Credit China provided by the third credit agency , to feed back to the bid evaluation experts whether the bidder and its legal representative and project leader have bribery crime records within the effective date stipulated in the bidding documents, whether the bidder has bad integrity records, etc., and the bid evaluation experts can make a decision after manual review Abandoned bid processing saves the time of bid evaluation experts, prevents human error, ensures fairness and justice in the process of verifying bidder information, and assists bid evaluation experts to conduct efficient bid evaluation.

As shown in Figure 2, the present invention also provides a system for verifying bidder information based on image recognition, including:

The obtaining module 1 is used to obtain the identity information picture of the bidder; wherein, the identity information picture includes a legal person certificate picture, a business license picture, a legal representative ID card picture, and a project leader ID card picture;

The preprocessing module 2 is used to preprocess the identity information picture of the bidder to obtain the corrected picture of the identity information of the bidder; wherein, the preprocessing includes tilt correction, grayscale, USM sharpening and two value;

The identification module 3 is used to identify the corrected image of the bidder's identity information to obtain the bidder's unified social credit code, the name of the legal representative and its citizen ID number, the name of the project leader and his citizen ID card Number;

The verification module 4 is used to obtain the valid date stipulated in the bidding documents, and when the current date is within the valid date, use a third-party credit reporting agency to verify the bidder and determine whether the bidder has a bad integrity record and bribery crime records; among them, the third-party credit reporting agency provides China Judgment Documents Network, Public Institution Online, and National Enterprise Credit Information Publicity System;

The first storage module 5 is used for extracting the uniform social credit code of the bidder, the name of the legal representative and its citizen ID number, the name of the project leader and its citizen ID number of the bidder when the verification fails, and Stored in the manual verification data table for manual verification;

The second storage module 6 stores the bidder's legal person certificate and business license into the verification pass data table when the verification is passed.

In one embodiment, also includes:

The storage module is used to obtain the image of the legal person certificate or business license of the institution/enterprise from the bidding documents and store it in the bidder qualification certificate image library, and obtain the ID card image of the legal representative from the bidding documents and store it in the identity of the legal representative ID card image library, obtain the ID card image of the project leader from the bidding documents and store it in the ID card image library of the project leader.

In one embodiment, the preprocessing module 2 includes:

A correction unit, configured to correct the inclination of the bidder's identity information picture to obtain a first processed picture when the bidder's identity information picture has a slope;

A grayscale processing unit, used to calculate the RGB value of each pixel in the identity information picture or the first processed picture when the bidder’s identity information picture has no inclination, or after obtaining the first processed picture Mapped to the grayscale value of 0-255 to get the grayscale processed image;

The calculation unit is used to record the segmentation threshold of the foreground and background of the gray scale processing picture as T, calculate the variance value, and make the segmentation threshold T traverse 0-255, calculate the variance value every time, and extract the square The T value with the largest difference is used as the target threshold;

A sharpening unit, configured to sharpen the grayscale processed picture by using the USM sharpening algorithm using the target threshold to obtain the sharpened processed picture;

The binarization unit is configured to perform binarization processing on the sharpened picture to obtain a corrected picture of the bidder's identity information.

In one embodiment, in the calculation unit, the calculation formula is:

为属于前景的像素点数占整幅图片的比例，其平均灰度为u₀，

为背景像素点数占整幅图片的比例，其平均灰度为u₁。Among them, g is the variance value, M×N is the size of the picture, N ₀ is the number of pixels whose grayscale value is less than the threshold T in the picture, N ₁ is the number of pixels in the picture whose grayscale is greater than the threshold T,

is the proportion of pixels belonging to the foreground to the whole picture, and its average gray level is u ₀ ,

is the proportion of the number of background pixels to the whole picture, and its average gray level is u ₁ .

In one embodiment, the binarization unit includes:

The horizontal scanning subunit is used to scan the sharpened picture horizontally, and add up the gray values of all points with light colors on both sides and dark colors in the middle of the sharpened picture as sum ₁ , and record the accumulated points as p ₁ ;

The vertical scanning sub-unit is used to vertically scan the sharpened image, and add up the gray values of all points whose upper and lower sides of the sharpened image are light in color and the middle is dark in color as sum ₂ , and record the accumulated points is p ₂ ;

The calculation subunit is used to calculate the threshold value Y of binarization; wherein the calculation formula is: Y=(sum ₁ +sum ₂ )/(p ₁ +p ₂ );

The binarization subunit is configured to use the threshold Y to binarize the sharpened picture to obtain a corrected picture of the bidder's identity information with the watermark removed.

In one embodiment, the recognition module 3 includes:

The first definition unit is used to use the character recognition network model to judge the area where the unified social credit code is located on the legal person certificate picture or business license picture of the institution/enterprise and define this area as a character area, and define other areas as text areas;

The second definition unit is used to identify the ID card picture of the legal representative and the ID card picture of the project leader by using the character recognition network model, judge the area where the citizen ID number is located and define this area as a character area, and define other areas as text areas ;

The first character recognition unit is used for character recognition on the character area of the legal person certificate picture or business license picture of the institution/enterprise to determine the legal representative of the bidder and the unified social credit code information;

The second character recognition unit is used for character recognition on the character area of the legal representative's ID card picture and the project leader's ID card picture, and determines the citizenship number of the legal representative of the bidder and the project leader;

The first text recognition unit is used to identify the text using OCR image recognition technology in the text area of the legal person certificate picture or business license picture of the institution/enterprise, and determine the name, category and legal representative of the bidder;

The second text recognition unit is used to identify the text using OCR image recognition technology in the text area of the legal representative's ID card picture and the project leader's ID card picture, and determine the name information of the legal representative of the bidder and the project leader.

In one embodiment, also includes:

The input module is used to automatically enter the legal person certificate or business license information of the public institution/enterprise that passed the verification in the data table into the bid evaluation system, and the image of the legal person certificate or business license of the public institution/enterprise that failed the manual verification The information is stored in the exception database.

The above-mentioned modules, units, and sub-units are all used to correspondingly execute each step in the above-mentioned method for verifying bidder information based on image recognition, and its specific implementation methods refer to the above-mentioned method embodiments, and will not be repeated here.

As shown in FIG. 3 , the present invention also provides a computer device, which may be a server, and its internal structure may be as shown in FIG. 3 . The computer device includes a processor, memory, network interface and database connected by a system bus. Among them, the processor designed by the computer is used to provide calculation and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, computer programs and databases. The memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database of the computer device is used to store all the data required in the process of the method of bidder information verification based on picture recognition. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by a processor, the method for verifying bidder information based on image recognition is realized.

Those skilled in the art can understand that the structure shown in FIG. 3 is only a block diagram of a partial structure related to the solution of this application, and does not constitute a limitation on the computer device on which the solution of this application is applied.

An embodiment of the present application also provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, any one of the above methods for verifying bidder information based on image recognition can be implemented.

Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above-mentioned embodiments can be completed by instructing related hardware through computer programs, and the computer programs can be stored in a non-volatile computer-readable memory In the medium, when the computer program is executed, it may include the processes of the embodiments of the above-mentioned methods. Wherein, any references to memory, storage, database or other media provided in the present application and used in the embodiments may include non-volatile and/or volatile memory. Nonvolatile memory can include read only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (SSRSDRAM), Enhanced SDRAM (ESDRAM), Synchronous Link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

Claims (10)

1. A method for bidder information verification based on image recognition, characterized in that the method comprises: Obtain the identity information picture of the bidder; the identity information picture includes the picture of the legal person certificate, the business license picture, the legal representative ID card picture and the project leader's ID card picture; Preprocess the bidder’s identity information image to obtain the corrected image of the bidder’s identity information; preprocessing includes tilt correction, grayscale, USM sharpening and binarization; Identify the corrected image of the bidder's identity information, and obtain the bidder's unified social credit code, legal representative name and citizen ID number, project leader's name and citizen ID number; Obtain the effective date stipulated in the bidding documents, and when the current date is within the effective date, use a third-party credit investigation agency to verify the bidder to determine whether the bidder has a record of bad faith and bribery; the third-party credit investigation agency Provide China Judgment Documents Network, Public Institution Online or National Enterprise Credit Information Publicity System; When the verification fails, extract the bidder's unified social credit code, legal representative's name and citizen ID number, project leader's name and citizen ID number, and store them in the manual verification data table for manual verification; When the verification is passed, the legal person certificate and business license of the bidder shall be stored in the verification data sheet. 2. A method for verifying bidder information based on picture identification according to claim 1, characterized in that before the step of obtaining the bidder's identity information picture, it also includes: Obtain images of legal person certificates or business licenses of public institutions, enterprises, institutions, and enterprises from bidding documents and store them in the bidder qualification certificate image library, obtain legal representative ID card images from bidding documents and store them in the legal representative ID card Image library, obtain the ID card picture of the project leader from the bidding documents and store it in the ID card image library of the project leader. 3. A method for verifying bidder information based on image recognition according to claim 1, characterized in that, the step of preprocessing the bidder's identity information image to obtain the bidder's identity information corrected image, include: When the bidder's identity information picture has an inclination, correct the inclination of the bidder's identity information picture to obtain a first processed picture; When the identity information picture of the bidder has no inclination or after obtaining the first processed picture, the RGB value of each pixel in the identity information picture or the first processed picture is mapped to a gray value of 0-255 to obtain gray processing pictures; Record the segmentation threshold of the foreground and background of the gray scale processing picture as T, calculate the variance value, and make the segmentation threshold T traverse 0-255, calculate the variance value each time, and extract the T value that maximizes the variance value as target threshold; Using the USM sharpening algorithm to sharpen the grayscale processed picture by using the target threshold to obtain the sharpened processed picture; Perform binarization on the sharpened image to obtain the corrected image of the bidder's identity information. 4. the method for a kind of bidder information verification based on picture identification according to claim 3, is characterized in that, the segmentation threshold value of the foreground and background of described gray-scale processing picture is denoted as T, the step of calculating variance value , the calculation formula is:

Among them, g is the variance value, M×N is the size of the picture, N ₀ is the number of pixels whose gray value of the pixel value in the picture is less than the threshold T, N ₁ is the number of pixels in the picture whose gray level is greater than the threshold T,

is the proportion of pixels belonging to the foreground to the whole picture, and the average gray level is u ₀ ;

is the proportion of the number of background pixels to the whole picture, and the average gray level is u ₁ . 5. A method for verifying bidder information based on picture recognition according to claim 3, wherein said sharpened picture is binarized to obtain the corrected picture of the bidder's identity information steps, including: Scan the sharpened image horizontally, and add up the gray values of all points with light colors on both sides and dark colors in the middle of the sharpened image as sum ₁ , and record the accumulated points as p ₁ ; Scanning the sharpened picture vertically, and summing up the gray values of all points whose upper and lower sides of the sharpened picture are light-colored and the middle is dark-colored are recorded as sum ₂ , and the accumulated points are recorded as p ₂ ; Calculate the threshold value Y of binarization; wherein the calculation formula is: Y=(sum ₁ +sum ₂ )/(p ₁ +p ₂ ); Using the threshold Y to binarize the sharpened picture to obtain a corrected picture of the bidder's identity information with the watermark removed. 6. the method for a kind of bidder's information verification based on picture identification according to claim 1, is characterized in that, the identity information correction picture of described bidder is identified, obtains the unified social credit code of described bidder, The steps for the name of the legal representative and his/her citizenship ID number, the name of the project leader and his/her citizenship ID number include: Use the character recognition network model to judge the area where the unified social credit code is located on the legal person certificate picture or business license picture of the institution, enterprise, institution, and enterprise, and define this area as a character area, and define other areas as text areas; Use the character recognition network model to identify the ID card picture of the legal representative and the ID card picture of the project leader, judge the area where the citizen ID number is located and define this area as a character area, and define other areas as text areas; Use character recognition for the character area of the legal person certificate picture or business license picture of the public institution, enterprise institution, enterprise, and determine the legal representative of the bidder and the unified social credit code information; Use character recognition on the character area of the legal representative's ID card picture and the project leader's ID card picture to determine the citizenship number of the legal representative of the bidder and the project leader; Use OCR image recognition technology to identify the text in the text area of the legal person certificate picture or business license picture of the institution, enterprise institution, enterprise, and determine the name, category and legal representative information of the bidder; Use OCR image recognition technology to identify the text in the text area of the legal representative's ID card picture and the project leader's ID card picture, and determine the name information of the legal representative of the bidder and the project leader. 7. A method for verifying bidder information based on image recognition according to claim 1, characterized in that, when the verification is passed, after the step of depositing the bidder's legal person certificate and business license into the verified data table, Also includes: Automatically enter the legal person certificate or business license information of public institutions, enterprises, institutions, and enterprises that pass the verification in the data sheet into the bid evaluation system, and automatically enter the legal person certificate pictures or business licenses of institutions, enterprises, institutions, and enterprises that fail manual verification The information is stored in the exception database. 8. A bidder information verification system based on image recognition, characterized in that it includes: The obtaining module is used to obtain the identity information picture of the bidder; wherein, the identity information picture includes a picture of a legal person certificate, a picture of a business license, a picture of the ID card of the legal representative, and a picture of the ID card of the project leader; The preprocessing module is used to preprocess the bidder's identity information picture to obtain the bidder's identity information corrected picture; wherein, the preprocessing includes tilt correction, grayscale, USM sharpening and binarization; The identification module is used to identify the corrected picture of the bidder's identity information, and obtain the bidder's unified social credit code, the name of the legal representative and the citizen's ID card number, the name of the project leader and the citizen's ID card number; The verification module is used to obtain the effective date stipulated in the bidding documents, and when the current date is within the effective date, use a third-party credit investigation agency to verify the bidder, and determine whether the bidder has a record of bad faith and bribery; Among them, the third-party credit reporting agency provides China Judgment Documents Network, Public Institution Online or the National Enterprise Credit Information Publicity System; The first storage module is used to extract the bidder's unified social credit code, legal representative name and citizen ID number, project leader's name and citizen ID number when the verification fails, and store them in the manual verification data table China and Israel for manual verification; The second storage module stores the bidder's legal person certificate and business license in the verification pass data table when the verification is passed. 9. A computer device, comprising a memory and a processor, the memory stores a computer program, wherein the processor implements the steps of the method according to any one of claims 1 to 7 when executing the computer program. 10. A computer-readable storage medium, on which a computer program is stored, characterized in that, when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images