WO2019024234A1 - Vehicle loss-related identification photo classification method and system, electronic device, and readable storage medium - Google Patents

Abstract

Disclosed in the present invention are a vehicle loss-related identification photo classification method and system, an electronic device, and a readable storage medium. The method comprises: receiving identity information and identification photos uploaded by vehicle insurance policyholders, identifying first photo category names corresponding to the identification photos using a preset type model, and generating first photo category name lists corresponding to the identity information, the first photo category name lists comprising the identified first photo category names, and the first photo category names being in mapping association with the corresponding identification photos; after receiving an identification photo calling instruction that carries the identity information of a vehicle insurance claimant transmitted by a claim terminal, searching for a first photo category name list corresponding to the identity information of the vehicle insurance claimant; and if a corresponding first photo category name list is found, feeding back the found first photo category name list to the claim terminal. By means of the present invention, the work efficiency of claim adjusters can be increased, and the accuracy is higher.

Description

Method, system, electronic device and readable storage medium for car damage certificate photo

Priority claim

This application claims the priority of the Chinese Patent Application entitled "Classification Method, System and Readable Storage Medium for Photographs of Vehicle Damage Documents" submitted on August 4, 2017, with the application number CN201710658108.3 submitted on August 4, 2017. The entire content of the application is incorporated herein by reference.

Technical field

The present invention relates to the field of computer technologies, and in particular, to a method, a system, an electronic device, and a readable storage medium for classifying a photo of a vehicle damage certificate.

Background technique

At present, in the field of auto insurance, there are a large number of photo documents of various types of auto insurance insured persons, such as ID card photos, bank card photos, and driver's license photos. When reviewing or reviewing claims, the claimant needs to verify the photos of these documents and compare them with the existing records in the system for risk control. However, a large number of photo IDs in the auto insurance claims system are often mixed and stored together, for example, photo of the reverse side of the ID card, photo of the front of the ID card, image of the reverse copy of the ID card, photo of the driver's license homepage, photo of the driver's license, and photo of the driver's license. The copy of the driver's license homepage image file is mixed together, and is not classified, so that the claimant needs to look for various photo IDs in many pictures for review or verification, which not only reduces the work efficiency of the claimants, but also has low accuracy. There is a risk of potential claims errors.

Summary of the invention

The main object of the present invention is to provide a method, a system, an electronic device and a readable storage medium for classifying a photo of a vehicle damage certificate, aiming at improving the work efficiency of the claimant and the accuracy of the car damage claim.

In order to achieve the above object, a first aspect of the present application provides an electronic device, where the electronic device includes a storage device and a processing device, and the storage device stores a classification system for a photo of a vehicle damage certificate that can be run on the processing device. When the classification system of the vehicle damage certificate photo is executed by the processing device, the following steps are implemented:

A. Receiving the identity information and the ID photo uploaded by the auto insurance applicant, using the trained preset type model to identify the first photo category name corresponding to each of the photo IDs, and generating a first photo category name corresponding to the identity information a list, the first photo category name list includes each of the identified first photo category names, and each of the first photo category names is associated with a corresponding ID photo map;

B. After receiving the document photo retrieval instruction issued by the claim terminal with the identity information of the auto insurance claimant, searching for a first photo category name list corresponding to the identity information of the auto insurance claimant;

C. If a corresponding first photo category name list is found, the found first photo category name list is fed back to the claim terminal, so that the claimant can retrieve each based on the found first photo category name list. Photo ID name associated with the mapped photo of the ID.

A second aspect of the present application provides a method for classifying a photo of a vehicle damage certificate, which is applied to an electronic device, and the method includes:

A. Receiving the identity information and the ID photo uploaded by the auto insurance applicant, using the trained preset type model to identify the first photo category name corresponding to each of the photo IDs, and generating a first photo category name corresponding to the identity information a list, the first photo category name list includes each of the identified first photo category names, and each of the first photo category names is associated with a corresponding ID photo map;

B. After receiving the document photo retrieval instruction issued by the claim terminal with the identity information of the auto insurance claimant, searching for a first photo category name list corresponding to the identity information of the auto insurance claimant;

C. If a corresponding first photo category name list is found, the found first photo category name list is fed back to the claim terminal, so that the claimant retrieves each photo category name association based on the found first photo category name list. Mapped photo of the ID.

A third aspect of the present application provides a classification system for a photo of a vehicle damage certificate, wherein the classification system of the vehicle damage certificate photo includes:

The identification module is configured to receive the identity information and the photo of the ID uploaded by the auto insurance applicant, and identify the first photo category name corresponding to each of the photo IDs by using the preset type model of the training, and generate a first corresponding to the identity information a list of photo category names, wherein the first photo category name list includes each of the identified first photo category names, and each of the first photo category names is associated with a corresponding document photo;

a search module, configured to search for a first photo category name list corresponding to the identity information of the auto insurance claimant after receiving the document photo retrieval instruction issued by the claim terminal with the identity information of the auto insurance claimant;

a feedback module, configured to: if the corresponding first photo category name list is found, feed back the found first photo category name list to the claim terminal, so that the claimant retrieves each photo based on the found first photo category name list The certificate name of the category name associated with the map.

A fourth aspect of the present application provides a computer readable storage medium having stored thereon at least one computer readable instruction executable by a processing device to:

Receiving the identity information and the ID photo uploaded by the auto insurance applicant, using the trained preset type model to identify the first photo category name corresponding to each of the photo IDs, and generating a first photo category name list corresponding to the identity information, The first photo category name list includes each of the identified first photo category names, and each of the first photo category names is associated with a corresponding certificate photo;

After receiving the document photo retrieval instruction issued by the claim terminal with the identity information of the auto insurance claimant, searching for a first photo category name list corresponding to the identity information of the auto insurance claimant;

If the corresponding first photo category name list is found, the found first photo category name list is fed back to the claim terminal, so that the claimant retrieves each photo category name association map based on the found first photo category name list. ID Photo.

The method, system, electronic device and readable storage medium for the photo of the vehicle damage certificate proposed by the invention are used to identify the auto insurance insured when the auto insurance policy insured Providing a first photo category name corresponding to each of the provided photo photos, and generating a first photo category name list corresponding to the identity information of the auto insurance policy holder, the first photo category name and the corresponding ID photo in the first photo category name list Correlation mapping; after the claim terminal issues the document photo retrieval instruction, the corresponding first photo category name list may be found according to the identity information of the auto insurance claimant, and the claimant can retrieve each photo category name based on the first photo category name list. Correlate the mapped photo of the ID without the need for the claimant to manually search for various ID photos in many pictures, and realize the automatic classification of the photo of the vehicle damage certificate, which not only improves the work efficiency of the claimant, but also has higher accuracy.

DRAWINGS

1 is a schematic diagram of an operating environment of a preferred embodiment of a classification system 10 for a vehicle damage certificate photograph of the present invention;

2 is a schematic diagram of functional modules of an embodiment of a classification system for a photo of a vehicle damage certificate according to the present invention;

3 is a schematic diagram of functional modules of another embodiment of a classification system for a photo of a vehicle damage certificate according to the present invention;

4 is a schematic flow chart of an embodiment of a method for classifying a photo of a vehicle damage certificate according to the present invention.

The implementation, functional features, and advantages of the present invention will be further described in conjunction with the embodiments.

Detailed ways

The present invention will be further described in detail below with reference to the accompanying drawings and embodiments, in order to make the present invention. It is understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

The invention provides a classification system for a photo of a vehicle damage certificate. Please refer to FIG. 1 , which is a schematic diagram of an operating environment of a preferred embodiment of the classification system 10 for a vehicle damage certificate photo of the present invention.

In the present embodiment, the classification system 10 of the vehicle damage certificate photograph is installed and operated in the electronic device 1. The electronic device 1 may include, but is not limited to, a storage device 11, a processing device 12, and a display 13. Figure 1 shows only the electronic device 1 with components 11-13, but it should be understood that not all illustrated components may be implemented, and more or fewer components may be implemented instead.

The storage device 11 is at least one type of readable computer storage medium, which in some embodiments may be an internal storage unit of the electronic device 1, such as a hard disk or memory of the electronic device 1. In other embodiments, the storage device 11 may also be an external storage device of the electronic device 1, such as a plug-in hard disk equipped on the electronic device 1, a smart memory card (Smart Media Card, SMC). ), Secure Digital (SD) card, Flash Card, etc. Further, the storage device 11 may also include both an internal storage unit of the electronic device 1 and an external storage device. The storage device 11 is configured to store application software installed on the electronic device 1 and various types of data, such as program codes of the classification system 10 of the vehicle damage certificate photo. The storage device 11 can also be used to temporarily store data that has been output or is about to be output.

The processing device 12 may be a central processing unit (CPU), a microprocessor or other data processing chip for running the storage device 11 in some embodiments. The stored program code or processing data, such as the classification system 10 that executes the photo of the vehicle damage certificate, and the like.

The display 13 in some embodiments may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch sensor, or the like. The display 13 is configured to display information processed in the electronic device 1 and a user interface for displaying visualization, such as identity information and ID photos sent by the insurance policy insured when insured, and the car insurance claimant issued by the claim terminal to be audited Identity information and ID photos, etc. The components 11-13 of the electronic device 1 communicate with one another via a system bus.

Please refer to FIG. 2 , which is a functional block diagram of a preferred embodiment of the classification system 10 for the vehicle damage certificate photo of the present invention. In this embodiment, the classification system 10 of the vehicle damage certificate photo may be divided into one or more modules, and the one or more modules are stored in the storage device 11 and are composed of one or more The processing device (this embodiment is the processing device 12) is executed to complete the present invention. For example, in FIG. 2, the classification system 10 of the vehicle damage certificate photo may be divided into an identification module 01, a search module 02, and a feedback module 03. The term "module" as used in the present invention refers to a series of computer program instruction segments capable of performing a specific function, which is more suitable than the program for describing the execution process of the classification system 10 of the vehicle-defective certificate photo in the electronic device 1. The following description will specifically describe the functions of the identification module 01, the search module 02, and the feedback module 03.

The identification module 01 is configured to receive the identity information and the ID photo uploaded by the auto insurance applicant, and identify the first photo category name corresponding to each of the photo IDs by using the trained preset type model, and generate a first corresponding to the identity information. a photo category name list, wherein the first photo category name list includes each of the identified first photo category names, and each of the first photo category names is associated with a corresponding ID photo.

In this embodiment, the classification system of the vehicle damage certificate photo can receive the identity information (for example, the ID number, etc.) and the ID photo of the auto insurance applicant issued by the auto insurance applicant (for example, the reverse photo of the ID card, the front photo of the ID card) , the image of the reverse copy of the ID card, the photo of the driver's license homepage, the photo of the driver's license, the photo of the driver's license, the photocopy of the driver's license, etc., for example, the receiving user is pre-arranged in the mobile phone, tablet, self-service terminal, etc. The insured request sent by the installed client, or the insured request sent by the user on the browser system in the terminal such as the mobile phone, the tablet computer, the self-service terminal device, such as the user may provide the interface provided by the client or the browser. After entering the identity information of the auto insurance policyholder and uploading the relevant photo of the ID card, the insurance system will issue a request for insurance to the classification system of the car damage certificate photo.

The classification system of the vehicle damage certificate photo identifies the first photo category name corresponding to each of the photo IDs by using the trained preset type model after receiving the insurance request with the auto insurance policy holder's identity information and the ID photo (for example, The reverse side of the ID card, the front of the ID card, the reverse copy of the ID card, the driver's license homepage, the driver's license page, the copy of the driver's license homepage, etc.). The preset type model can be continuously trained, learned, verified, optimized, etc. by continuously labeling sample pictures of a plurality of different photo categories and identifying the sample pictures of different photo categories, so as to train them accurately. Identify models of different photo category names. For example, the preset type model may adopt a Convolutional Neural Network (CNN) model or the like.

Identifying the first photo category corresponding to each of the photo IDs by using the trained preset type model After the name, a first photo category name list corresponding to the identity information of the auto insurance policy holder may be generated, where the first photo category name list includes each of the identified first photo category names, wherein the first photo category name Each of the first photo category names in the list is associated with the corresponding ID photo, that is, the first photo category name list has a mapping association relationship between each first photo category name and the corresponding ID photo. For example, in the first photo category name list, a corresponding photo retrieving option is generated after each of the first photo category names, and the corresponding photo photo can be retrieved by clicking the photo retrieving option; or Each of the first photo category name display areas includes a link address of the corresponding ID photo, and the first photo category name is clicked to directly retrieve the photo of the ID corresponding to the clicked first photo category name.

The searching module 02 is configured to search for the first photo corresponding to the identity information of the auto claimant after receiving the document photo retrieval instruction issued by the claim terminal with the identity information of the auto claimant (eg, the ID card number, etc.) A list of photo category names.

The feedback module 03 is configured to: if the corresponding first photo category name list is found, feed back the found first photo category name list to the claim terminal, so that the claimant can retrieve each of the first photo category name lists based on the found first photo category name list. The photo category name is associated with the mapped photo of the ID. For example, in the first photo category name list, each of the first photo category names is generated with a corresponding photo retrieval option, and after the found first photo category name list is fed back to the claim terminal, the claimant only needs to The corresponding photo of the photo can be retrieved by clicking the photo retrieval option corresponding to the photo category name in the first photo category name list according to the photo category name currently to be retrieved. Alternatively, each of the first photo category name display areas includes a link address of a corresponding ID photo, and the claimant can automatically retrieve the location by simply clicking on the photo category name currently to be retrieved in the first photo category name list. The photo of the ID corresponding to the photo category name of the click.

In the embodiment, when the automobile insurance policy insured is insured, the first type of photo category corresponding to each photo of the document provided by the auto insurance applicant is identified by using the preset type model of the training, and the first corresponding to the identity information of the auto insurance insured is generated. a list of photo category names, each first photo category name in the first photo category name list is associated with a corresponding photo ID; after the claim terminal issues a photo photo retrieval instruction, the corresponding information may be found according to the identity information of the auto insurance claimant A photo category name list, the claimant can retrieve the photo of the photo map associated with each photo category name based on the first photo category name list, without the claimant manually searching for various ID photos in a plurality of pictures, and realizing the vehicle damage certificate The automatic classification of photos not only improves the efficiency of claimants, but also has higher accuracy and reduces the risk of claims errors.

Preferably, the feedback module 03 is further configured to:

If the selection instruction of the photo category name in the found first photo category name list sent by the claim terminal is received, the photo of the photo associated with the photo category name corresponding to the selection instruction is retrieved, and the photo will be adjusted. The taken photo of the ID is fed back to the claim terminal. For example, in the first photo category name list, each of the first photo category names is generated with a corresponding photo retrieving option, and clicking the photo retrieving option is to issue a selection instruction to the corresponding photo category name; Or each of the stated The first photo category name display area includes a link address of the corresponding ID photo, and clicking the first photo category name is to issue a selection instruction to the first photo category name. After receiving the selection instruction of the photo category name in the found first photo category name list sent by the claim terminal, the ID photo associated with the selected photo category name can be found, and the found photo of the ID is found. The photo of the retrieved document is fed back to the claim terminal.

As shown in FIG. 3, another embodiment of the present invention provides a classification system for a photo of a vehicle damage certificate. Based on the foregoing embodiment, the method further includes a matching association module 04, where:

The identification module 01 is further configured to:

Receiving identity information (for example, ID number) and ID photo uploaded by the car insurance claimant (for example, photo of the reverse side of the ID card, photo of the front of the ID card, image of the reverse copy of the ID card, photo of the driver's license homepage, photo of the driver's license page, The license card homepage copy image file, etc., uses the trained preset type model to identify the second photo category name corresponding to each of the photo IDs.

The searching module 02 is further configured to:

Find a list of first photo category names corresponding to the identity information.

The matching association module 04 is further configured to:

If the corresponding first photo category name list is found, the first photo category name in the first photo category name list is matched with the identified second photo category name, for example, if there is the identified second photo category If the name is the same as a first photo category name, the second photo category name is associated with the first photo category name, or if the identified second photo category name and a first photo category name are preset Associated category (this associated category can be a more closely related photo category, such as the front photo of the ID card and the reverse photo of the ID card, the photo of the driver's license home page and the photo of the driver's license, etc.), then the second photo category name Associated with the first photo category name. Generating, according to the association result, a second photo category name list, where the second photo category name list includes a first photo category name in the first photo category name list, and the identified second photo category name, each of the first The photo category name is associated with the corresponding certificate photo, and each of the second photo category names is associated with the corresponding ID photo.

Further, the first photo category name and the second photo category name of the matching association are placed in the same display line or display column display. In an embodiment, in the second photo category name list, each of the first photo category names is followed by a corresponding photo retrieval option, and the photo retrieval option is clicked to retrieve the corresponding a photo of the certificate, or each of the first photo category name display areas includes a link address of the corresponding photo ID, and clicking the first photo category name may retrieve the photo of the ID corresponding to the clicked first photo category name; After the second photo category name is generated, a corresponding photo retrieving option is generated, and the corresponding photo ID can be retrieved by clicking the photo retrieving option, or each of the second photo category name display areas has a corresponding The link address of the ID photo, click on the second photo category name, and the photo of the ID corresponding to the clicked second photo category name can be retrieved. In another implementation manner, in the second photo category name list, the first photo category name and the second photo category name that match the association correspond to one photo retrieving option, and the photo retrieving option may be clicked to retrieve Match the associated ID photo and ID photo, or match the associated first photo The display line or display column of the slice category name and the second photo category name, including the link address of the corresponding ID photo, click on the display row area or the display column area, and the corresponding ID photo and ID photo can be retrieved.

In this embodiment, when the claim is made, the second photo category name of the ID photo sent by the claimant may be automatically classified, and the photo in the first photo category name list generated by the claimant when insured may be used. The category names are matched to generate a second photo category name list. The staff on the claim side can retrieve the photo of the photo of each photo category sent by the claimant and the photo of the photo of each photo category sent by the claimant according to the second photo category name list, so as to provide the same photo category when the claim is made. The photo of the ID is the same as the photo of the ID provided during the insured, so as to carry out the claim review, no need to manually find all kinds of ID photos in many pictures, which not only improves the work efficiency of the claimants, but also has higher accuracy.

Further, in other embodiments, the preset type model is a Regions of Convolutional Neural Network (RCNN) model, and the training process of the preset type model is as follows:

S1, preparing a preset number (for example, 1000 sheets) of photo samples of the photo with the corresponding photo category name for each preset photo category;

S2. Dividing the photo samples corresponding to each preset photo category into a first proportional (eg, 70%) training subset and a second proportional (eg, 30%) verification subset, and each training subset is The photo samples are mixed to obtain a training set, and the photo samples in each verification subset are mixed to obtain a verification set;

S3. Train the preset type model by using the training set;

S4. Verify, by using the verification set, the accuracy of the preset type model of the training. If the accuracy rate is greater than or equal to the preset accuracy rate, the training ends, or if the accuracy rate is less than the preset accuracy rate, increase each one. Predetermining the number of the certificate photo samples corresponding to the photo category, and performing steps S2, S3, and S4 again until the accuracy of the trained preset type model is greater than or equal to the preset accuracy.

As shown in Table 1 below, in a specific embodiment, there are a total of 18 preset photo categories (the reverse side of the ID card, the front of the ID card, the reverse copy of the ID card, etc.), and the corresponding category is set for each photo category name. No., according to the order of the category number, identify and train the photo samples of each preset photo category until the model training of all preset photo categories is completed, and finally obtain a volume that can accurately identify the 18 preset photo categories. Neural network regional model.

Photo category name	Category number
Reverse side of ID card	1
ID card front	2
Copy of the reverse side of the ID card	3
Copy of the front of the ID card	4
Driver's license homepage	5
Driver's license page	6
Copy of the driver's license homepage	7

Copy of the driver's license page	8
The first page of the driving permit	9
The second page of the driving license	10
Driving license page 3	11
Copy of the first page of the driving license	12
Copy of the second page of the driving license	13
Copy of the third page of the driving license	14
Bank card	15
Copy of bank card	16
Warranty	17
Copy of policy	18

Table 1

The invention further provides a method for classifying photos of vehicle damage certificates.

Referring to FIG. 4, FIG. 4 is a schematic flow chart of an embodiment of a method for classifying a photo of a vehicle damage certificate according to the present invention.

In an embodiment, the method for classifying the photo of the vehicle damage certificate comprises:

Step S10, receiving the identity information and the ID photo uploaded by the auto insurance policy holder, using the trained preset type model to identify the first photo category name corresponding to each of the photo IDs, and generating a first photo category corresponding to the identity information. a name list, wherein the first photo category name list includes each of the identified first photo category names, and each of the first photo category names is associated with a corresponding ID photo.

In this embodiment, the classification system of the vehicle damage certificate photo can receive the identity information (for example, the ID number, etc.) and the ID photo of the auto insurance applicant issued by the auto insurance applicant (for example, the reverse photo of the ID card, the front photo of the ID card) , the image of the reverse copy of the ID card, the photo of the driver's license homepage, the photo of the driver's license, the photo of the driver's license, the photocopy of the driver's license, etc., for example, the receiving user is pre-arranged in the mobile phone, tablet, self-service terminal, etc. The insured request sent by the installed client, or the insured request sent by the user on the browser system in the terminal such as the mobile phone, the tablet computer, the self-service terminal device, such as the user may provide the interface provided by the client or the browser. After entering the identity information of the auto insurance policyholder and uploading the relevant photo of the ID card, the insurance system will issue a request for insurance to the classification system of the car damage certificate photo.

The classification system of the vehicle damage certificate photo identifies the first photo category name corresponding to each of the photo IDs by using the trained preset type model after receiving the insurance request with the auto insurance policy holder's identity information and the ID photo (for example, The reverse side of the ID card, the front of the ID card, the reverse copy of the ID card, the driver's license homepage, the driver's license page, the copy of the driver's license homepage, etc.). The preset type model can be continuously trained, learned, verified, optimized, etc. by continuously labeling sample pictures of a plurality of different photo categories and identifying the sample pictures of different photo categories, so as to train them accurately. Identify models of different photo category names. For example, the preset type model may adopt a Convolutional Neural Network (CNN) model or the like.

After identifying the first photo category name corresponding to each of the photo IDs by using the preset type model of the training, a first photo category name list corresponding to the identity information of the auto insurance applicant may be generated, where the first photo category name list is Including the identified first photo category names, wherein each first photo category name in the first photo category name list is associated with a corresponding ID photo, that is, the first photo category name list is in each A relationship between a photo category name and a corresponding document photo is established. For example, in the first photo category name list, a corresponding photo retrieving option is generated after each of the first photo category names, and the corresponding photo photo can be retrieved by clicking the photo retrieving option; or Each of the first photo category name display areas includes a link address of the corresponding ID photo, and the first photo category name is clicked to directly retrieve the photo of the ID corresponding to the clicked first photo category name.

Step S20, after receiving the document photo retrieval instruction issued by the claim terminal with the identity information of the auto insurance claimant (for example, the ID card number, etc.), searching for the first photo category name corresponding to the identity information of the auto insurance claimant List.

Step S30: If the corresponding first photo category name list is found, the found first photo category name list is fed back to the claim terminal, so that the claimant retrieves each photo category name based on the found first photo category name list. Associate the mapped photo of the ID. For example, in the first photo category name list, each of the first photo category names is generated with a corresponding photo retrieval option, and after the found first photo category name list is fed back to the claim terminal, the claimant only needs to The corresponding photo of the photo can be retrieved by clicking the photo retrieval option corresponding to the photo category name in the first photo category name list according to the photo category name currently to be retrieved. Alternatively, each of the first photo category name display areas includes a link address of a corresponding ID photo, and the claimant can automatically retrieve the location by simply clicking on the photo category name currently to be retrieved in the first photo category name list. The photo of the ID corresponding to the photo category name of the click.

In the embodiment, when the automobile insurance policy insured is insured, the first type of photo category corresponding to each photo of the document provided by the auto insurance applicant is identified by using the preset type model of the training, and the first corresponding to the identity information of the auto insurance insured is generated. a list of photo category names, each first photo category name in the first photo category name list is associated with a corresponding photo ID; after the claim terminal issues a photo photo retrieval instruction, the corresponding information may be found according to the identity information of the auto insurance claimant A photo category name list, the claimant can retrieve the photo of the photo map associated with each photo category name based on the first photo category name list, without the claimant manually searching for various ID photos in a plurality of pictures, and realizing the vehicle damage certificate The automatic classification of photos not only improves the efficiency of claimants, but also has higher accuracy and reduces the risk of claims errors.

Further, in other embodiments, the method further includes:

If the selection instruction of the photo category name in the found first photo category name list sent by the claim terminal is received, the photo of the photo associated with the photo category name corresponding to the selection instruction is retrieved, and the photo will be adjusted. The taken photo of the ID is fed back to the claim terminal. For example, in the first photo category name list, each of the first photo category names is generated with a corresponding photo retrieval option, and the click The photo retrieval option is to issue a selection instruction to the corresponding photo category name; or each of the first photo category name display areas includes a link address of the corresponding ID photo, and clicking the first photo category name is The first photo category name issues a selection instruction. After receiving the selection instruction of the photo category name in the found first photo category name list sent by the claim terminal, the ID photo associated with the selected photo category name can be found, and the found photo of the ID is found. The photo of the retrieved document is fed back to the claim terminal.

Further, in other embodiments, after the step S10, the method further includes:

Receive identity information (for example, ID number) and ID photo of the car insurance claimant (for example, photo of the reverse side of the ID card, photo of the front of the ID card, image of the reverse copy of the ID card, photo of the driver's license homepage, photo of the driver's license page, driving The homepage copy image file, etc., uses the trained preset type model to identify the second photo category name corresponding to each of the photo IDs.

Find a list of first photo category names corresponding to the identity information.

If the corresponding first photo category name list is found, the first photo category name in the first photo category name list is matched with the identified second photo category name, for example, if there is the identified second photo category If the name is the same as a first photo category name, the second photo category name is associated with the first photo category name, or if the identified second photo category name and a first photo category name are preset Associated category (this associated category can be a more closely related photo category, such as the front photo of the ID card and the reverse photo of the ID card, the photo of the driver's license home page and the photo of the driver's license, etc.), then the second photo category name Associated with the first photo category name. Generating, according to the association result, a second photo category name list, where the second photo category name list includes a first photo category name in the first photo category name list, and the identified second photo category name, each of the first The photo category name is associated with the corresponding certificate photo, and each of the second photo category names is associated with the corresponding ID photo.

Further, the first photo category name and the second photo category name of the matching association are placed in the same display line or display column display. In an embodiment, in the second photo category name list, each of the first photo category names is followed by a corresponding photo retrieval option, and the photo retrieval option is clicked to retrieve the corresponding a photo of the certificate, or each of the first photo category name display areas includes a link address of the corresponding photo ID, and clicking the first photo category name may retrieve the photo of the ID corresponding to the clicked first photo category name; After the second photo category name is generated, a corresponding photo retrieving option is generated, and the corresponding photo ID can be retrieved by clicking the photo retrieving option, or each of the second photo category name display areas has a corresponding The link address of the ID photo, click on the second photo category name, and the photo of the ID corresponding to the clicked second photo category name can be retrieved. In another implementation manner, in the second photo category name list, the first photo category name and the second photo category name that match the association correspond to one photo retrieving option, and the photo retrieving option may be clicked to retrieve Matching the associated ID photo and ID photo, or matching the first photo category name and the second photo category name display line or display column, including the link address of the corresponding ID photo, clicking the display line area or the display column Area, you can retrieve the corresponding ID photo and ID photo.

In this embodiment, when the claim is made, the second photo category name of the ID photo sent by the claimant may be automatically classified, and the photo in the first photo category name list generated by the claimant when insured may be used. The category names are matched to generate a second photo category name list. The staff on the claim side can retrieve the photo of the photo of each photo category sent by the claimant and the photo of the photo of each photo category sent by the claimant according to the second photo category name list, so as to provide the same photo category when the claim is made. The photo of the ID is the same as the photo of the ID provided during the insured, so as to carry out the claim review, no need to manually find all kinds of ID photos in many pictures, which not only improves the work efficiency of the claimants, but also has higher accuracy.

Further, in other embodiments, the preset type model is a Regions of Convolutional Neural Network (RCNN) model, and the training process of the preset type model is as follows:

S1, preparing a preset number (for example, 1000 sheets) of photo samples of the photo with the corresponding photo category name for each preset photo category;

S2. Dividing the photo samples corresponding to each preset photo category into a first proportional (eg, 70%) training subset and a second proportional (eg, 30%) verification subset, and each training subset is The photo samples are mixed to obtain a training set, and the photo samples in each verification subset are mixed to obtain a verification set;

S3. Train the preset type model by using the training set;

S4. Verify, by using the verification set, the accuracy of the preset type model of the training. If the accuracy rate is greater than or equal to the preset accuracy rate, the training ends, or if the accuracy rate is less than the preset accuracy rate, increase each one. Predetermining the number of the certificate photo samples corresponding to the photo category, and performing steps S2, S3, and S4 again until the accuracy of the trained preset type model is greater than or equal to the preset accuracy.

As shown in Table 1 below, in a specific embodiment, there are a total of 18 preset photo categories (the reverse side of the ID card, the front of the ID card, the reverse copy of the ID card, etc.), and the corresponding category is set for each photo category name. No., according to the order of the category number, identify and train the photo samples of each preset photo category until the model training of all preset photo categories is completed, and finally obtain a volume that can accurately identify the 18 preset photo categories. Neural network regional model.

Photo category name	Category number
Reverse side of ID card	1
ID card front	2
Copy of the reverse side of the ID card	3
Copy of the front of the ID card	4
Driver's license homepage	5
Driver's license page	6
Copy of the driver's license homepage	7
Copy of the driver's license page	8
The first page of the driving permit	9

The second page of the driving license	10
Driving license page 3	11
Copy of the first page of the driving license	12
Copy of the second page of the driving license	13
Copy of the third page of the driving license	14
Bank card	15
Copy of bank card	16
Warranty	17
Copy of policy	18

Table 1

Furthermore, the present invention also provides a computer readable storage medium storing a classification system of a vehicle damage certificate photograph, the classification system of the vehicle damage certificate photograph being executable by at least one processing device such that The at least one processing device performs the steps of the method for classifying the vehicle damage certificate photo in the above embodiment, and the specific implementation processes of the steps S10, S20, and S30 of the method for classifying the vehicle damage certificate photo are as described above, and are not Let me repeat.

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 can also be implemented by hardware, but in many cases, the former is A better implementation. Based on such understanding, the technical solution of the present invention, 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, disk, The optical disc includes a number of instructions for causing a terminal device (which may be a cell phone, a computer, a server, an air conditioner, or a network device, etc.) to perform the methods described in various embodiments of the present invention.

The preferred embodiments of the present invention have been described above with reference to the drawings, and are not intended to limit the scope of the invention. The serial numbers of the embodiments of the present invention are merely for the description, and do not represent the advantages and disadvantages of the embodiments. Additionally, although logical sequences are shown in the flowcharts, in some cases the steps shown or described may be performed in a different order than the ones described herein.

A person skilled in the art can implement the invention in various variants without departing from the scope and spirit of the invention. For example, the features of one embodiment can be used in another embodiment to obtain a further embodiment. Any modifications, equivalent substitutions and improvements made within the technical concept of the invention are intended to be included within the scope of the invention.

Claims (20)

1. An electronic device, comprising: a storage device and a processing device, wherein the storage device stores a classification system for a photo of a vehicle damage certificate that can be run on the processing device, the vehicle damage certificate photo The classification system is implemented by the processing device to implement the following steps:

   A. Receiving the identity information and the ID photo uploaded by the auto insurance applicant, using the trained preset type model to identify the first photo category name corresponding to each of the photo IDs, and generating a first photo category name corresponding to the identity information a list, the first photo category name list includes each of the identified first photo category names, and each of the first photo category names is associated with a corresponding ID photo map;

   B. After receiving the document photo retrieval instruction issued by the claim terminal with the identity information of the auto insurance claimant, searching for a first photo category name list corresponding to the identity information of the auto insurance claimant;

   C. If a corresponding first photo category name list is found, the found first photo category name list is fed back to the claim terminal, so that the claimant retrieves each photo category name association based on the found first photo category name list. Mapped photo of the ID.
2. The electronic device according to claim 1, wherein the processing device is further configured to execute a classification system of the vehicle damage certificate photo to implement the following steps:

   If the selection instruction of the photo category name in the found first photo category name list sent by the claim terminal is received, the photo of the photo associated with the photo category name corresponding to the selection instruction is retrieved, and the photo will be adjusted. The taken photo of the ID is fed back to the claim terminal.
3. The electronic device according to claim 1, wherein after the step A, the processing device is further configured to execute a classification system of the vehicle damage certificate photo to implement the following steps:

   Receiving the identity information and the ID photo uploaded by the auto insurance applicant, and using the trained preset type model to identify the second photo category name corresponding to each of the photo photos;

   Find a list of first photo category names corresponding to the identity information;

   If the corresponding first photo category name list is found, the first photo category name in the first photo category name list is matched with the identified second photo category name, and the second photo category name list is generated based on the association result. The second photo category name list includes a first photo category name in the first photo category name list, and the identified second photo category name, and each of the second photo category names and the corresponding ID photo association map .
4. The electronic device according to claim 3, wherein the processing device is further configured to execute a classification system of the vehicle damage certificate photo to implement the following steps:

   The first photo category name and the second photo category name of the matching association are placed in the same display line or display column display.
5. The electronic device according to any one of claims 1 to 4, wherein the preset type model is a convolutional neural network region model, and the training process of the preset type model is as follows:

   S1, preparing a preset number of photo samples of the photo with the corresponding photo category name for each preset photo category;

   S2. Dividing the photo samples corresponding to each preset photo category into a first proportional training subset and a second proportional verification subset, mixing the photo samples in each training subset to obtain a training set, and Verify that the photo samples in the subset are mixed to get a validation set;

   S3. Train the preset type model by using the training set;

   S4. Verify, by using the verification set, the accuracy of the preset type model of the training. If the accuracy rate is greater than or equal to the preset accuracy rate, the training ends, or if the accuracy rate is less than the preset accuracy rate, increase each one. The number of the certificate photo samples corresponding to the preset photo category is preset, and steps S2, S3, and S4 are re-executed.
6. A method for classifying a photo of a vehicle damage certificate is applied to an electronic device, characterized in that the method comprises:

   A. Receiving the identity information and the ID photo uploaded by the auto insurance applicant, using the trained preset type model to identify the first photo category name corresponding to each of the photo IDs, and generating a first photo category name corresponding to the identity information a list, the first photo category name list includes each of the identified first photo category names, and each of the first photo category names is associated with a corresponding ID photo map;

   B. After receiving the document photo retrieval instruction issued by the claim terminal with the identity information of the auto insurance claimant, searching for a first photo category name list corresponding to the identity information of the auto insurance claimant;

   C. If a corresponding first photo category name list is found, the found first photo category name list is fed back to the claim terminal, so that the claimant retrieves each photo category name association based on the found first photo category name list. Mapped photo of the ID.
7. The method for classifying a photo of a vehicle damage certificate according to claim 6, wherein the method further comprises:

   If the selection instruction of the photo category name in the found first photo category name list sent by the claim terminal is received, the photo of the photo associated with the photo category name corresponding to the selection instruction is retrieved, and the photo will be adjusted. The taken photo of the ID is fed back to the claim terminal.
8. The method for classifying a photo of a vehicle damage document according to claim 6, wherein after the step A, the method further comprises:

   Receiving the identity information and the ID photo uploaded by the auto insurance applicant, and using the trained preset type model to identify the second photo category name corresponding to each of the photo photos;

   Find a list of first photo category names corresponding to the identity information;

   If the corresponding first photo category name list is found, the first photo category name in the first photo category name list is matched with the identified second photo category name, and the second photo category name list is generated based on the association result. The second photo category name list includes a first photo category name in the first photo category name list, and the identified second photo category name, and each of the second photo category names and the corresponding ID photo association map .
9. The method for classifying a photo of a vehicle damage certificate according to claim 8, further comprising:

   The first photo category name and the second photo category name of the matching association are placed in the same display line or display column display.
10. The method for classifying a photo of a vehicle damage certificate according to any one of claims 6-9, wherein the preset type model is a convolutional neural network region model, and the training process of the preset type model is as follows:

    S1, preparing a preset number of photo samples of the photo with the corresponding photo category name for each preset photo category;

    S2. Dividing the photo samples corresponding to each preset photo category into a first proportional training subset and a second proportional verification subset, mixing the photo samples in each training subset to obtain a training set, and Verify that the photo samples in the subset are mixed to get a validation set;

    S3. Train the preset type model by using the training set;

    S4. Verify, by using the verification set, the accuracy of the preset type model of the training. If the accuracy rate is greater than or equal to the preset accuracy rate, the training ends, or if the accuracy rate is less than the preset accuracy rate, increase each one. The number of the certificate photo samples corresponding to the preset photo category is preset, and steps S2, S3, and S4 are re-executed.
11. A classification system for a photo of a vehicle damage certificate, characterized in that the classification system of the photo of the vehicle damage certificate comprises:

    The identification module is configured to receive the identity information and the photo of the ID uploaded by the auto insurance applicant, and identify the first photo category name corresponding to each of the photo IDs by using the preset type model of the training, and generate a first corresponding to the identity information a list of photo category names, wherein the first photo category name list includes each of the identified first photo category names, and each of the first photo category names is associated with a corresponding document photo;

    a search module, configured to search for a first photo category name list corresponding to the identity information of the auto insurance claimant after receiving the document photo retrieval instruction issued by the claim terminal with the identity information of the auto insurance claimant;

    a feedback module, configured to: if the corresponding first photo category name list is found, feed back the found first photo category name list to the claim terminal, so that the claimant retrieves each photo based on the found first photo category name list The certificate name of the category name associated with the map.
12. The classification system for a vehicle damage document photo according to claim 11, wherein the feedback module is further configured to:

    If the selection instruction of the photo category name in the found first photo category name list sent by the claim terminal is received, the photo of the photo associated with the photo category name corresponding to the selection instruction is retrieved, and the photo will be adjusted. The taken photo of the ID is fed back to the claim terminal.
13. A classification system for a vehicle damage certificate photo according to claim 11, further comprising a matching association module, wherein:

    The identification module is further configured to:

    Receiving the identity information and the ID photo uploaded by the auto insurance applicant, and using the trained preset type model to identify the second photo category name corresponding to each of the photo photos;

    The lookup module is further configured to:

    Find a list of first photo category names corresponding to the identity information;

    The matching association module is used to:

    If the corresponding first photo category name list is found, the first photo category name in the first photo category name list is matched with the identified second photo category name, and the second photo category name list is generated based on the association result. The second photo category name list includes a first photo category name in the first photo category name list, and the identified second photo category name, and each of the second photo category names and the corresponding ID photo association map .
14. The classification system for a vehicle damage certificate photo according to claim 13, wherein the matching association module is further configured to:

    The first photo category name and the second photo category name of the matching association are placed in the same display line or display column display.
15. The classification system of the vehicle damage certificate photo according to any one of claims 11 to 14, wherein the preset type model is a convolutional neural network region model, and the training process of the preset type model is as follows:

    S1, preparing a preset number of photo samples of the photo with the corresponding photo category name for each preset photo category;

    S2. Dividing the photo samples corresponding to each preset photo category into a first proportional training subset and a second proportional verification subset, mixing the photo samples in each training subset to obtain a training set, and Verify that the photo samples in the subset are mixed to get a validation set;

    S3. Train the preset type model by using the training set;

    S4. Verify, by using the verification set, the accuracy of the preset type model of the training. If the accuracy rate is greater than or equal to the preset accuracy rate, the training ends, or if the accuracy rate is less than the preset accuracy rate, increase each one. The number of the certificate photo samples corresponding to the preset photo category is preset, and steps S2, S3, and S4 are re-executed.
16. A computer readable storage medium having stored thereon at least one computer readable instruction executable by a processing device to:

    Receiving the identity information and the ID photo uploaded by the auto insurance applicant, using the trained preset type model to identify the first photo category name corresponding to each of the photo IDs, and generating a first photo category name list corresponding to the identity information, The first photo category name list includes each of the identified first photo category names, and each of the first photo category names is associated with a corresponding certificate photo;

    After receiving the document photo retrieval instruction issued by the claim terminal with the identity information of the auto insurance claimant, searching for a first photo category name list corresponding to the identity information of the auto insurance claimant;

    If the corresponding first photo category name list is found, the found first photo category name list is fed back to the claim terminal, so that the claimant retrieves each photo category name association map based on the found first photo category name list. ID Photo.
17. The computer readable storage medium of claim 16 further comprising:

    If the selection instruction of the photo category name in the found first photo category name list sent by the claim terminal is received, the photo of the photo associated with the photo category name corresponding to the selection instruction is retrieved, and the photo will be adjusted. The taken photo of the ID is fed back to the claim terminal.
18. The computer readable storage medium of claim 16 further comprising:

    Receiving the identity information and the ID photo uploaded by the auto insurance applicant, and using the trained preset type model to identify the second photo category name corresponding to each of the photo photos;

    Find a list of first photo category names corresponding to the identity information;

    If the corresponding first photo category name list is found, the first photo category name in the first photo category name list is matched with the identified second photo category name, and the second photo category name list is generated based on the association result. The second photo category name list includes a first photo category name in the first photo category name list, and the identified second photo category name, and each of the second photo category names and the corresponding ID photo association map .
19. The computer readable storage medium of claim 18, further comprising:

    The first photo category name and the second photo category name of the matching association are placed in the same display line or display column display.
20. The computer readable storage medium according to any one of claims 16 to 19, wherein the preset type model is a convolutional neural network region model, and the training process of the preset type model is as follows:

    S1, preparing a preset number of photo samples of the photo with the corresponding photo category name for each preset photo category;

    S2. Dividing the photo samples corresponding to each preset photo category into a first proportional training subset and a second proportional verification subset, mixing the photo samples in each training subset to obtain a training set, and Verify that the photo samples in the subset are mixed to get a validation set;

    S3. Train the preset type model by using the training set;

    S4. Verify, by using the verification set, the accuracy of the preset type model of the training. If the accuracy rate is greater than or equal to the preset accuracy rate, the training ends, or if the accuracy rate is less than the preset accuracy rate, increase each one. The number of the certificate photo samples corresponding to the preset photo category is preset, and steps S2, S3, and S4 are re-executed.

Images