WO2019214308A1 - 一种理赔业务的数据处理方法、装置、设备及服务器 - Google Patents

一种理赔业务的数据处理方法、装置、设备及服务器 Download PDF

Info

Publication number
WO2019214308A1
WO2019214308A1 PCT/CN2019/075435 CN2019075435W WO2019214308A1 WO 2019214308 A1 WO2019214308 A1 WO 2019214308A1 CN 2019075435 W CN2019075435 W CN 2019075435W WO 2019214308 A1 WO2019214308 A1 WO 2019214308A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
request
feature data
similarity
historical
Prior art date
Application number
PCT/CN2019/075435
Other languages
English (en)
French (fr)
Inventor
胡越
Original Assignee
阿里巴巴集团控股有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 阿里巴巴集团控股有限公司 filed Critical 阿里巴巴集团控股有限公司
Publication of WO2019214308A1 publication Critical patent/WO2019214308A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/22—Matching criteria, e.g. proximity measures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/08—Insurance

Definitions

  • the embodiment of the present specification belongs to the technical field of computer data processing for insurance anti-fraud identification, and in particular, to a data processing method, device, device and server for a claims service.
  • Insurance is the financial and personal protection that can be enjoyed by paying the prescribed premiums. With the economic development of society and the awareness of people's insurance, the demand for insurance business is also increasing. In order to provide users with faster claims services, many insurance companies gradually began to support self-determined claims. In this way, users can only take and upload the pictures needed for claims, and insurance company staff can complete claims without on-site audit. deal with. For example, when the claims for auto insurance self-photographing, the claims for Internet e-commerce transaction insurance, and the claims for broken screen insurance are processed, the claim maker can upload pictures by means of the claim, and the insurance company judges whether it is dangerous based on the picture content.
  • the embodiment of the present specification aims to provide a data processing method, device, device and server for a claims service, which can automatically determine whether a claim image submitted by a user has a picture of a case image that has been claimed for fraudulent use, and thus more directly and effectively recognizes Fraud, and can greatly reduce the cost of manual review, improve recognition efficiency and accuracy.
  • a data processing method for a claims business comprising:
  • the fraudulent image in the claim image is determined based on the degree of similarity.
  • a data processing device for a claims business comprising:
  • a feature processing module configured to calculate claim feature data of the claim image corresponding to the claim request
  • a similarity calculation module configured to calculate a degree of similarity between the claim feature data and feature data of a historical similar image, where the historical homogeneous image includes image information included in a claim type of the same request claim type corresponding to the claim request;
  • the piracy identification module is configured to determine a pirate image in the claim image based on the degree of similarity.
  • a data processing device for a claims service comprising a processor and a memory for storing processor executable instructions, the processor implementing the instructions to:
  • the fraudulent image in the claim image is determined based on the degree of similarity.
  • a server comprising at least one processor and a memory for storing processor-executable instructions, the processor implementing the instructions to:
  • the fraudulent image in the claim image is determined based on the degree of similarity.
  • the data processing method, device, device and server of the claim service provided by the embodiment of the present specification perform feature extraction on the claim image uploaded by the claim claimant, and judge the similarity degree of the claim image and the image of the same type of claim compensation. Whether the claim image is stolen from the case image of the history has been claimed.
  • the embodiment of the present specification can calculate the similarity between the application claim image and the historical image library of the similar case in an offline or real-time manner, and automatically, effectively and efficiently discover the fraudulent behavior of the fraudulent historical image for claim settlement, improve the recognition processing efficiency and the accuracy of the recognition, and reduce the manual. Review costs.
  • FIG. 1 is a schematic flow chart of an embodiment of a data processing method for a claims service provided by the present specification
  • FIG. 2 is a schematic flow chart of another embodiment of a data processing method for the claims service provided by the present specification
  • FIG. 3 is a schematic flow chart of another embodiment of a data processing method for the claims service provided by the present specification.
  • FIG. 4 is a schematic flow chart of another embodiment of a data processing method for the claims service provided by the present specification.
  • FIG. 5 is a block diagram showing a hardware structure of a server for processing claims data processing according to an embodiment of the present invention
  • FIG. 6 is a block diagram showing the structure of an embodiment of a data processing apparatus for a claims service provided by the present specification
  • FIG. 7 is a block diagram showing the structure of an embodiment of a data processing apparatus of another claim service provided by the present specification.
  • FIG. 8 is a block diagram showing the structure of an embodiment of a data processing apparatus for another claims service provided by the present specification.
  • the agent of the insurance fraud may use the image of the case in which the claim was previously made as the image basis for claiming the case in his own accident case.
  • These historical case images that have been previously processed may be images that the perpetrator used to make claims, or may be images that others have used to claim.
  • the image information of the same type of claim case is usually stolen.
  • the agent may steal the accident scene photo of the car insurance claim, and may steal the image of other suit damage claims when the suit is in the suit.
  • the embodiment of the present specification may compare the feature of the claim image with the image of the historical same type claim case. If there is an image in the history gallery that is similar to the claim image and exceeds the threshold, the claim may be decided.
  • the image has the risk of stealing pictures, thus effectively identifying the behavior of fraudulent fraud in the settlement of claims by stealing historical images.
  • FIG. 1 is a schematic flowchart diagram of an embodiment of a data processing method for the claim service provided by the present specification.
  • the present specification provides method operation steps or device structures as shown in the following embodiments or figures, there may be more or partial merged fewer operational steps in the method or device based on conventional or no inventive labor. Or module unit.
  • the execution order of the steps or the module structure of the device is not limited to the execution order or the module structure shown in the embodiment or the drawings.
  • server or terminal product of the method or module structure When the device, server or terminal product of the method or module structure is applied, it may be executed sequentially or in parallel according to the method or module structure shown in the embodiment or the drawing (for example, parallel processor or multi-thread processing). Environment, even including distributed processing, server cluster implementation environment).
  • the description of the embodiment of the auto insurance insurance business claim fraud identification provided below does not constitute a limitation on other scalable application scenarios based on the present specification.
  • the embodiments provided in this specification can also be applied to an implementation scenario of image fraud identification of a transaction dispute, evidence image of a virtual currency transaction claim, and the like, such as a shopping platform or a merchant identifying an item uploaded by a buyer return request. Whether there is a stolen image in the damaged image.
  • the descriptions of the corresponding terms such as “claim request” and “claim image” may be correspondingly described according to different implementation scenarios, such as The return request and the damaged item image in the above example may correspond to the claim request and the claim image in the solution of the embodiment of the present specification.
  • the following is an example of the applicable scenario of the implementation of the auto insurance insurance business claim fraud.
  • the description of the implementation of the other implementation scenarios is described with reference to the description of the implementation process in this embodiment, and is not described herein.
  • the data processing method of the claim service provided by the present specification may include:
  • the user can initiate a claim request through the terminal.
  • the claim image may be uploaded together when the claim is initiated, such as capturing a damaged image of the vehicle through the mobile terminal or selecting an image from the terminal's gallery application as the claim image.
  • the claim may be initiated first, and the claim image is uploaded after the claim is accepted.
  • the image described in the embodiment of the present specification may include a picture in multiple formats such as a bitmap, a vector diagram, and the like, and may also include an image captured in the video.
  • the image may also include a video, which may be viewed as a collection of consecutive images.
  • the server performing the claims service processing may perform processing in real time, or may perform storage first, and then perform image data processing based on the trigger instruction.
  • the server may calculate the claim feature data of the claim image by using one or more preset algorithms.
  • the corresponding claim feature data may be calculated for each claim image uploaded by the user.
  • algorithm models can be used to extract the feature vector of the image, and the feature vector is compared with the feature vector of the historical similar image.
  • SIFT Scale-invariant Feature Transform
  • SURF Speeded Up Robust Features
  • HOG Histogram of Oriented Gradient
  • other feature data such as RGB (red, green, blue, and three primary colors), gradient, grayscale, and the like, or other custom algorithms may be used to extract the feature data of the image.
  • a learning network of a machine learning CNN may be used to extract image feature vectors.
  • CNN is a machine learning model under deep supervision and learning. It has strong adaptability, is good at mining local features of data, and extracts global training features and classifications.
  • CNN is used to extract the feature data of the image, and the image features can be extracted more efficiently and accurately, and a better recognition result is obtained.
  • the network model adopted by the specific CNN network may include a variant of the CNN network or an improved machine learning network model, and the model network layer construction adopted in the specific implementation process may be correspondingly set according to the scenario requirements or design requirements, and will not be described herein. .
  • the CNN may be used to extract the image feature vector, or other methods, or CNN and other methods may be used. Combine to achieve the extraction of image feature data.
  • S2 Calculate the degree of similarity between the claim feature data and the feature data of the historical similar image, the historical similar image including the image information included in the claim type of the request claim type corresponding to the claim request.
  • the calculation of the feature data of the historical similar image may also be obtained by extracting the parameters in the above manner.
  • the calculation of the feature data of the historical similar image and the calculation of the feature data of the claim image may use the same algorithm, but it is not excluded that other feature calculation algorithms may be used in other embodiments, and then the feature data is Conversion/transformation/mapping, etc., to calculate the degree of similarity between the claim image and the historical image of the same type.
  • the feature data of the claim image (herein collectively referred to as claim feature data) is compared with the feature data of the historical similar image to determine the degree of similarity between the two.
  • the historical homogeneous image may generally include image information included in a claim type of the same request claim type corresponding to the claim request. If the claim type corresponding to the claim request is auto insurance, the historical similar image may include a historical image in which the insurance type is also a car insurance. These historical images may include images of cases in which claims have been previously made, images of cases in which claims processing is being performed, and even image information obtained from Internet/partners and the like.
  • the insurance company can accumulate image data of the claim case and store it in the photo library.
  • the images of the photo library may be stored in the storage blocks of the same claim type according to the types of claims marked as different claim types or the same claim type.
  • the feature data calculation of the historical similar image may be performed for image extraction and calculation during the processing of the claims business, or may be calculated in advance, and the feature data is calculated when the image is stored in the image library, and the feature data is stored together or separately. In this way, when performing the similarity calculation of the claim request, the feature data of the pre-computed historical similar image can be directly retrieved, and the calculation speed is accelerated.
  • the server may calculate feature data corresponding to the image in the same type of claim case based on the claim type corresponding to the claim request, and then perform similarity calculation on the claim feature data of the claim image and the feature data of the historical similar image.
  • the similarity calculation may specifically design corresponding calculation manners, such as comparison of feature vector values, difference calculation, feature data association degree, and the like.
  • the degree of similarity calculated may be expressed by a score, or may be represented by other feature symbols, such as dividing the similarity levels AAA, AA, A, BBB, BB, and the like.
  • S4 Determine a theft image in the claim image based on the degree of similarity.
  • a corresponding judgment characterization may be set. For example, when a score threshold is set, when the similarity score of the claim image and a historical similar image exceeds a threshold, the claim image may be determined as a pirate image.
  • the specific threshold setting can be set according to the actual application scenario or processing requirements.
  • the setting may also be determined in combination with other data information, such as image generation or acquisition time information, image source, or marking of other images, Attribute information, etc.
  • image generation or acquisition time information such as image generation or acquisition time information, image source, or marking of other images, Attribute information, etc.
  • multiple thresholds may be set, and different thresholds correspond to different levels of image theft/risk level, for example, the degree of similarity is more than 90%, and the similarity is between 60% and 76%.
  • the image is a level 1 misappropriation, similar to 50%-60% for second-level theft.
  • the risk identification, monitoring, and the like can be further performed based on the stolen image.
  • the data processing method of the claim service provided by the embodiment of the present specification performs feature extraction on the claim image uploaded by the claim claimant, and compares the claim image with the historical image of the same type of claim to determine the degree of misappropriation of the claim image.
  • the embodiment of the present specification can calculate the similarity between the application claim image and the historical image library of the similar case in an offline or real-time manner, and automatically, effectively and efficiently discover the fraudulent behavior of the fraudulent historical image for claim settlement, improve the recognition processing efficiency and the accuracy of the recognition, and reduce the manual. Review costs.
  • the present specification further provides another embodiment of the method, as described in FIG. 2, and specifically, the method further includes:
  • S6 Determine a fraud identification result of the claim request based on the stolen image.
  • how to determine whether the claim request has fraudulent behavior according to the stolen image may perform setting of the corresponding rule.
  • a claim application scenario such as insurance
  • the fraud claim may be determined to be fraudulent. Therefore, in another embodiment of the method, the determining, by the fraudulent image, the fraud identification result of the claim request may include:
  • the number of the stolen images may be further combined to determine whether the claim request has fraudulent behavior.
  • the number of claims images collected in some application scenarios is very large. Compared with the claims in previous historical claims cases (such as the same vehicle model in a car accident, similar collision angle, similar accident scene environment, etc.), there may be actual claims.
  • the image is similar to the image in the historical claims case. Therefore, in this embodiment, the number of the stolen images identified may be further combined to determine whether fraudulent behavior exists. For example, if the claim image has a large number of similarities with the image of the historical claim case, and exceeds the set number decision threshold, the risk identification result of the fraud claim may be outputted, which can be well applied in similar claims cases and effectively recognized. Fraudulent use of images. Therefore, in another embodiment of the method provided by the present specification, the determining the fraud identification result of the claim request based on the fraudulent image comprises:
  • the subject matter of the images involved in different types of claims cases will not or hardly overlap or resemble, and the image subjects such as mobile insurance and car damage insurance are usually significantly different.
  • the types of claims of similar products/subjects of similar products may be associated in advance, and if no stolen images are found in the same claim type, the search range may be further expanded. Searching in the associated claims type gallery to find out if there are images in the other claims type.
  • the method further includes:
  • S82 Determine a theft image in the claim image based on the obtained degree of similarity of the association.
  • the specific types of images as the associated claim types can be set according to the actual application scenarios and design requirements, and can refer to the classification of physical products or services (types of insurance, etc.), or custom classifications, or primary classification, secondary A multi-level classification such as classification performs corresponding relationship setting.
  • the associated claim type and the claim type corresponding to the claim image do not belong to the same type, and the similarity decision threshold may be adjusted correspondingly when determining whether the similarity is similar, such as in the associated claim type.
  • the similarity decision threshold may be lower than the similarity decision threshold in the same type, at which point multiple pirated images may be determined.
  • the two can also set the same value, or when a stricter decision mechanism is adopted based on the lower-order data range, the similarity decision threshold in the associated claim type is higher than the similarity decision threshold in the same type.
  • the fraudulent images determined by different image search ranges may be set with different fraud behavior decision mechanisms, and the number of the stolen images determined from the associated images of the associated types may be collected to determine whether the claims are fraudulent. behavior.
  • the method may further include:
  • S10 Determine a fraud identification result of the claim request based on the number of the stolen images determined from the related image.
  • a fraudulent image may be determined from the associated image, that is, the fraud claim is determined, and other implementation manners may be set, such as the number of stolen images determined by the same historical image and the fraud detected in the associated image. The number of images is jointly determined to determine whether there is fraud.
  • the dashed lines in Figure 4 indicate that other embodiments may be performed in conjunction with one or more processing steps in other embodiments.
  • the data processing method of the claim service provided by the embodiment of the present specification performs feature extraction on the claim image uploaded by the claim claimant, and compares the claim image with the historical image of the same type of claim to determine the degree of misappropriation of the claim image.
  • the embodiment of the present specification can calculate the similarity between the application claim image and the historical image library of the similar case in an offline or real-time manner, and automatically, effectively and efficiently discover the fraudulent behavior of the fraudulent historical image for claim settlement, improve the recognition processing efficiency and the accuracy of the recognition, and reduce the manual. Review costs.
  • the risk prompt may be further performed or the corresponding claim processing flow may be automatically performed. Make sure that you don't use the stolen image, you can complete the subsequent claims process, such as automatic claims or manual review. When it is found that the historical claim image is stolen, the claim can be rejected, or the risk warning or warning light of the claim proof is not true.
  • the method described above can be used for the identification of insurance fraud on the server side, such as a server of an insurance company or other service platform, and can also be applied to other terminal devices, such as a PC (personal computer) machine, a server, an industrial computer (industrial). Control computer), mobile smart phone, tablet electronic device, portable computer (such as laptop computer, etc.), personal digital assistant (PDA), or desktop computer or smart wearable device.
  • a PC personal computer
  • Control computer mobile smart phone
  • tablet electronic device portable computer (such as laptop computer, etc.)
  • PDA personal digital assistant
  • desktop computer or smart wearable device Mobile communication terminal, handheld device, in-vehicle device, wearable device, television device, distributed or integrated computing device.
  • the system server may include a separate server, a server cluster, a distributed system server, or a server that processes device request data and other associated data processing, as may be applied to a system server of an insurance business or a servant or a third party.
  • System server combination may be used for the identification of insurance fraud on the server side
  • FIG. 5 is a hardware structural block diagram of a server to which a claim service data processing is applied according to an embodiment of the present invention.
  • server 10 may include one or more (only one of which is shown) processor 100 (processor 100 may include, but is not limited to, a processing device such as a microprocessor MCU or a programmable logic device FPGA), A memory 200 for storing data, and a transmission module 300 for communication functions. It will be understood by those skilled in the art that the structure shown in FIG.
  • server 10 may also include more or fewer components than those shown in FIG. 5, such as other processing hardware, such as a database or multi-level cache, GPU (image processor), or with FIG. Show different configurations.
  • processing hardware such as a database or multi-level cache, GPU (image processor), or with FIG. Show different configurations.
  • the memory 200 can be used to store software programs and modules of application software, such as program instructions/modules corresponding to the search method in the embodiment of the present invention, and the processor 100 executes various functions by running software programs and modules stored in the memory 200.
  • Application and data processing that is, a processing method for realizing the content display of the above navigation interaction interface.
  • Memory 200 can include high speed random access memory, and can also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid state memory.
  • memory 200 can further include memory remotely located relative to processor 100, which can be connected to computer terminal 10 over a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.
  • the transmission module 300 is configured to receive or transmit data via a network.
  • the network specific examples described above may include a wireless network provided by a communication provider of the computer terminal 10.
  • the transmission module 300 includes a Network Interface Controller (NIC) that can be connected to other network devices through a base station to communicate with the Internet.
  • the transmission module 300 can be a Radio Frequency (RF) module for communicating with the Internet wirelessly.
  • NIC Network Interface Controller
  • RF Radio Frequency
  • the present specification also provides a data processing apparatus for a claims service.
  • the apparatus may include a system (including a distributed system), software (applications), modules, components, servers, clients, etc., using the methods described in the embodiments of the present specification, in conjunction with necessary device hardware for implementing the hardware.
  • the processing device in one embodiment provided by this specification is as described in the following embodiments.
  • the apparatus described in the following embodiments is preferably implemented in software, hardware, or a combination of software and hardware, is also possible and contemplated.
  • FIG. 6 is a schematic structural diagram of a module of a data processing apparatus of a claim service provided by the present disclosure, which may include:
  • the feature processing module 101 is configured to calculate claim feature data of the claim image corresponding to the claim request;
  • the similarity calculation module 102 may be configured to calculate a degree of similarity between the claim feature data and feature data of a historical similar image, where the historical homogeneous image includes an image included in a claim type having the same request claim type corresponding to the claim request information;
  • the fraud identification module 103 can be configured to determine a theft image in the claim image based on the degree of similarity.
  • the apparatus may further include:
  • the fraud identification module 104 can be configured to determine a fraud identification result of the claim request based on the stolen image.
  • the fraud identification module 104 may determine whether the user's claim request has fraudulent behavior according to the set rule when the fraud identification module 103 identifies the stolen image.
  • the fraud identification module 104 may include:
  • the first identifying unit 1041 may be configured to determine that the claim request has fraudulent behavior if it is determined that at least one fraudulent image exists.
  • the fraud identification module 104 can include:
  • the second identifying unit may be configured to determine that the claim request has fraudulent behavior if the number of the stolen images reaches a quantity decision threshold.
  • another embodiment of the device may further include:
  • the neighborhood search module 105 may be configured to calculate an association similarity between the claim feature data and the feature data of the associated image, where the associated image includes the determined image information included in the claim type associated with the request claim type ;
  • the association and theft identification module 106 may be configured to determine the fraudulent image in the claim image based on the obtained degree of similarity of the association.
  • the associated theft identification module 106 can include:
  • the number identification processing unit 1061 may be configured to determine a fraud identification result of the claim request based on the number of the stolen images determined from the associated image.
  • the server or the client provided by the embodiment of the present specification may be implemented by a processor executing a corresponding program instruction in a computer, such as a C++ language of a Windows operating system, implemented on a PC side or a server side, or other systems such as Linux, android/iOS.
  • a processor executing a corresponding program instruction in a computer, such as a C++ language of a Windows operating system, implemented on a PC side or a server side, or other systems such as Linux, android/iOS.
  • the present specification also provides a data processing device for a claims service, and may specifically include a processor and a memory for storing processor-executable instructions, the processor implementing the instructions to:
  • the fraudulent image in the claim image is determined based on the degree of similarity.
  • the processor may also implement other processing based on the description of the foregoing embodiment, such as determining a fraud identification result of the claim request based on the stolen image, and the like.
  • the description of the specific processing device is described with reference to the foregoing corresponding method embodiments, and details are not described herein.
  • the above instructions may be stored in a variety of computer readable storage media.
  • the computer readable storage medium may include physical means for storing information, which may be digitized and stored in a medium utilizing electrical, magnetic or optical means.
  • the computer readable storage medium of this embodiment may include: means for storing information by means of electrical energy, such as various types of memories, such as RAM, ROM, etc.; means for storing information by magnetic energy means, such as hard disk, floppy disk, magnetic tape, magnetic Core memory, bubble memory, U disk; means for optically storing information such as CD or DVD.
  • electrical energy such as various types of memories, such as RAM, ROM, etc.
  • magnetic energy means such as hard disk, floppy disk, magnetic tape, magnetic Core memory, bubble memory, U disk
  • means for optically storing information such as CD or DVD.
  • quantum memories graphene memories, and the like.
  • the processing device may specifically provide a server for anti-fraud identification of the claims service for the insurance server or the third-party service organization, and the server may be a server, a server cluster, a distributed system server, or a server that requests data from the processing device.
  • the server may be a server, a server cluster, a distributed system server, or a server that requests data from the processing device.
  • embodiments of the present specification also provide a specific server product, the server including at least one processor and a memory for storing processor-executable instructions, the processor implementing the instructions to:
  • the fraudulent image in the claim image is determined based on the degree of similarity.
  • the apparatus, the processing device, and the server described in the foregoing embodiments of the present specification may further include other embodiments according to the description of the related method embodiments.
  • the apparatus, the processing device, and the server described in the foregoing embodiments of the present specification may further include other embodiments according to the description of the related method embodiments.
  • the feature data calculation of the claim image may be performed in advance.
  • Build/process implementation This specification does not exclude the implementation of online computing/storage, etc. in other embodiments.
  • historical image extraction, feature calculation, image library accumulation, fraud identification, and the like can be performed online in real time.
  • the data processing method, device, device and server of the claim service provided by the embodiment of the present specification perform feature extraction on the claim image uploaded by the claim claimant, and judge the similarity degree of the claim image and the image of the same type of claim compensation. Whether the claim image is stolen from the case image of the history has been claimed.
  • the embodiment of the present specification can calculate the similarity between the application claim image and the historical image library of the similar case in an offline or real-time manner, and automatically, effectively and efficiently discover the fraudulent behavior of the fraudulent historical image for claim settlement, improve the recognition processing efficiency and the accuracy of the recognition, and reduce the manual. Review costs.
  • the embodiments of the present specification refer to operations such as image feature extraction, similarity threshold setting, image or feature data existence, and the like, data acquisition, storage, interaction, calculation, judgment, and the like, and the data description
  • the embodiments of the present specification are It is not limited to situations that must be consistent with industry communication standards, standard CNN network models or feature-lifting algorithms, communication protocols, and standard data models/templates or embodiments of the specification.
  • Certain industry standards or implementations that have been modified in a manner that uses a custom approach or an embodiment described above may also achieve the same, equivalent, or similar, or post-deformation implementation effects of the above-described embodiments.
  • Embodiments obtained by applying such modified or modified data acquisition, storage, judgment, processing, etc. may still fall within the scope of alternative embodiments of the present specification.
  • PLD Programmable Logic Device
  • FPGA Field Programmable Gate Array
  • HDL Hardware Description Language
  • the controller can be implemented in any suitable manner, for example, the controller can take the form of, for example, a microprocessor or processor and a computer readable medium storing computer readable program code (eg, software or firmware) executable by the (micro)processor.
  • computer readable program code eg, software or firmware
  • examples of controllers include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, The Microchip PIC18F26K20 and the Silicone Labs C8051F320, the memory controller can also be implemented as part of the memory's control logic.
  • the controller can be logically programmed by means of logic gates, switches, ASICs, programmable logic controllers, and embedding.
  • Such a controller can therefore be considered a hardware component, and the means for implementing various functions included therein can also be considered as a structure within the hardware component.
  • a device for implementing various functions can be considered as a software module that can be both a method of implementation and a structure within a hardware component.
  • the processing device, device, module or unit set forth in the above embodiments may be implemented by a computer chip or an entity, or by a product having a certain function.
  • a typical implementation device is a computer.
  • the computer can be, for example, a personal computer, a laptop computer, a car-mounted human-machine interaction device, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet.
  • a computer, wearable device, or a combination of any of these devices are examples of these devices.
  • the above devices are described as being separately divided into various modules by function.
  • the functions of the modules may be implemented in the same software or software, or the modules that implement the same function may be implemented by multiple sub-modules or a combination of sub-units.
  • the device embodiments described above are merely illustrative.
  • the division of the unit is only a logical function division.
  • there may be another division manner for example, multiple units or components may be combined or integrated. Go to another system, or some features can be ignored or not executed.
  • the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, and may be in an electrical, mechanical or other form.
  • the controller can be logically programmed by means of logic gates, switches, ASICs, programmable logic controllers, and embedding.
  • the computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture comprising the instruction device.
  • the apparatus implements the functions specified in one or more blocks of a flow or a flow and/or block diagram of the flowchart.
  • These computer program instructions can also be loaded onto a computer or other programmable data processing device such that a series of operational steps are performed on a computer or other programmable device to produce computer-implemented processing for execution on a computer or other programmable device.
  • the instructions provide steps for implementing the functions specified in one or more of the flow or in a block or blocks of a flow diagram.
  • a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
  • processors CPUs
  • input/output interfaces network interfaces
  • memory volatile and non-volatile memory
  • the memory may include non-persistent memory, random access memory (RAM), and/or non-volatile memory in a computer readable medium, such as read only memory (ROM) or flash memory.
  • RAM random access memory
  • ROM read only memory
  • Memory is an example of a computer readable medium.
  • Computer readable media includes both permanent and non-persistent, removable and non-removable media.
  • Information storage can be implemented by any method or technology.
  • the information can be computer readable instructions, data structures, modules of programs, or other data.
  • Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read only memory. (ROM), electrically erasable programmable read only memory (EEPROM), flash memory or other memory technology, compact disk read only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, Magnetic tape cartridges, magnetic tape storage or other magnetic storage devices or any other non-transportable media can be used to store information that can be accessed by a computing device.
  • computer readable media does not include temporary storage of computer readable media, such as modulated data signals and carrier waves.
  • embodiments of the present specification can be provided as a method, system, or computer program product.
  • embodiments of the present specification can take the form of an entirely hardware embodiment, an entirely software embodiment or a combination of software and hardware.
  • embodiments of the present specification can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) including computer usable program code.
  • Embodiments of the present description can be described in the general context of computer-executable instructions executed by a computer, such as a program module.
  • program modules include routines, programs, objects, components, data structures, and the like that perform particular tasks or implement particular abstract data types.
  • Embodiments of the present specification can also be practiced in distributed computing environments where tasks are performed by remote processing devices that are connected through a communication network.
  • program modules can be located in both local and remote computer storage media including storage devices.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Physics & Mathematics (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • Marketing (AREA)
  • Artificial Intelligence (AREA)
  • General Business, Economics & Management (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • Development Economics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Technology Law (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Image Analysis (AREA)
  • Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)

Abstract

本说明书实施例公开了一种理赔业务的数据处理方法、装置、设备及服务器，通过对理赔请求人员上传的理赔图像进行特征提取，将理赔图像与历史同类型理赔的图像进行相似程度判断，判断出理赔图像是否盗用历史已经理赔的案件图像。本说明书实施例可以线下或实时计算申请理赔图片与同类案件历史图片库相似度，自动、有效、高效的发现盗用历史图像进行理赔的欺诈行为，提高识别处理效率和识别的准确性，降低人工审核成本。

Description

一种理赔业务的数据处理方法、装置、设备及服务器 技术领域
本说明书实施例方案属于保险反欺诈识别的计算机数据处理的技术领域，尤其涉及一种理赔业务的数据处理方法、装置、设备及服务器。
背景技术
保险是通过缴纳规定的保费，然后可以享受的财务、人身等保障。随着社会的经济发展和人们保险意识的提高，保险业务的需求也越来越多。为了给用户提供更快的理赔服务，许多保险公司逐渐开始支持自主理赔，这种方式下，用户可以只需要拍摄并上传理赔所需的图片，保险公司工作人员可以无需现场审核，就可以完成理赔处理。例如车险自助拍照的理赔、互联网电商交易保障保险的理赔、碎屏险的理赔等理赔业务处理时，可以通过理赔发起者上传图片的方式，保险公司基于图片内容判断是否出险。
然而，由于保险理赔有一定的经济杠杆效应，使得市场上出现大量骗保、欺诈的行为，对保险行业的健康发展带来非常不利的影响，损坏保险公司和公众利益。目前传统的来识别欺诈行为的方式还是主要依靠人工逐个对保单的信息审核、对用户上传图像的肉眼识别。但目前随着保险业务量的不断增大，尤其是碎片化、低保费的互联网保险种类越来越多，人工审核的工作量非常大，人工理赔的成本很高。并且，由于人工理赔存在主观、业务能力、经验等方面的限制，保险欺诈识别率也较为低下，一旦发生保险欺诈，对用户体验和公司形象、财产等造成较大损失。
因此，业内亟需一种可以更加有效和高效的识别出保险欺诈的处理方式。
发明内容
本说明书实施例目的在于提供一种理赔业务的数据处理方法、装置、设备及服务器，可以自动判断用户提交的理赔图像是否存在盗用历史已经理赔的案件图像的图片，进而更加直接、有效的识别出欺诈行为，并可以大幅度降低人工审核成本，提高识别效率和准确性。
本说明书实施例提供的一种理赔业务的数据处理方法、装置、设备及服务器是包括以下方式实现的：
一种理赔业务的数据处理方法，所述方法包括：
计算理赔请求对应的理赔图像的理赔特征数据；
计算所述理赔特征数据与历史同类图像的特征数据的相似程度，所述历史同类图像包括与所述理赔请求对应的请求理赔类型相同的理赔类型中所包含的图像信息；
基于得到相似程度确定所述理赔图像中的盗用图像。
一种理赔业务的数据处理装置，包括：
特征处理模块，用于计算理赔请求对应的理赔图像的理赔特征数据；
相似计算模块，用于计算所述理赔特征数据与历史同类图像的特征数据的相似程度，所述历史同类图像包括与所述理赔请求对应的请求理赔类型相同的理赔类型中所包含的图像信息；
盗用识别模块，用于基于得到相似程度确定所述理赔图像中的盗用图像。
一种理赔业务的数据处理设备，包括处理器以及用于存储处理器可执行指令的存储器，所述处理器执行所述指令时实现：
计算理赔请求对应的理赔图像的理赔特征数据；
计算所述理赔特征数据与历史同类图像的特征数据的相似程度，所述历史同类图像包括与所述理赔请求对应的请求理赔类型相同的理赔类型中所包含的图像信息；
基于得到相似程度确定所述理赔图像中的盗用图像。
一种服务器，包括至少一个处理器以及用于存储处理器可执行指令的存储器，所述处理器执行所述指令时实现：
计算理赔请求对应的理赔图像的理赔特征数据；
计算所述理赔特征数据与历史同类图像的特征数据的相似程度，所述历史同类图像包括与所述理赔请求对应的请求理赔类型相同的理赔类型中所包含的图像信息；
基于得到相似程度确定所述理赔图像中的盗用图像。
本说明书实施例提供的一种理赔业务的数据处理方法、装置、设备及服务器，通过对理赔请求人员上传的理赔图像进行特征提取，将理赔图像与历史同类型理赔的图像进行相似程度判断，判断出理赔图像是否盗用历史已经理赔的案件图像。本说明书实施例可以线下或实时计算申请理赔图片与同类案件历史图片库相似度，自动、有效、高效的 发现盗用历史图像进行理赔的欺诈行为，提高识别处理效率和识别的准确性，降低人工审核成本。
附图说明
为了更清楚地说明本说明书实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本说明书中记载的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。
图1是本说明书提供的一种理赔业务的数据处理方法实施例的流程示意图；
图2是本说明书提供的另一种理赔业务的数据处理方法实施例的流程示意图；
图3是本说明书提供的另一种理赔业务的数据处理方法实施例的流程示意图；
图4是本说明书提供的另一种理赔业务的数据处理方法实施例的流程示意图；
图5是应用本发明实施例的一种理赔业务数据处理的服务器的硬件结构框图；
图6是本说明书提供的一种理赔业务的数据处理装置实施例的模块结构示意图；
图7是本说明书提供的另一种理赔业务的数据处理装置实施例的模块结构示意图；
图8是本说明书提供的另一种理赔业务的数据处理装置实施例的模块结构示意图。
具体实施方式
为了使本技术领域的人员更好地理解本说明书中的技术方案，下面将结合本说明书实施例中的附图，对本说明书实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本说明书中的一部分实施例，而不是全部的实施例。基于本说明书中的一个或多个实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都应当属于本说明书实施例保护的范围。
在一些保险欺诈的案件中，由于事故案件未真实发生，或事实为全部/部分伪造，保险欺诈的行为人可能会使用以前进行理赔的案件图像作为自己事故案件进行理赔的图像依据。这些以前进行过理赔的历史案件图像可能是行为人自己出现理赔使用过的图像，也可能是其他人理赔使用过的图像。一般的，欺诈的行为人在某一类理赔案件进行请求理赔时，若使用历史理赔图像，则通常会盗取同类型理赔案件的图像信息。例如， 在汽车理赔案件中，行为人可能会盗用汽车车险理赔的事故现场照片，在西服服装类理赔时，可能会盗用其他西服破损理赔的图像等等。因此，本说明书实施例在接收用户理赔请求时，可以对将理赔图像与历史同类型理赔案件的图像进行特征对比，若历史图库中存在与理赔图像相似程度超过阈值的图像，则可以判决该理赔图像存在盗图的风险，从而有效识别理赔中通过盗用历史图像进行保险欺诈，骗取保费的行为。
下面以一个具体的保险业务欺诈识别处理的应用场景为例对本说明书实施方案进行说明。具体的，图1是本说明书提供的所述一种理赔业务的数据处理方法实施例的流程示意图。虽然本说明书提供了如下述实施例或附图所示的方法操作步骤或装置结构，但基于常规或者无需创造性的劳动在所述方法或装置中可以包括更多或者部分合并后更少的操作步骤或模块单元。在逻辑性上不存在必要因果关系的步骤或结构中，这些步骤的执行顺序或装置的模块结构不限于本说明书实施例或附图所示的执行顺序或模块结构。所述的方法或模块结构的在实际中的装置、服务器或终端产品应用时，可以按照实施例或者附图所示的方法或模块结构进行顺序执行或者并行执行(例如并行处理器或者多线程处理的环境、甚至包括分布式处理、服务器集群的实施环境)。
当然，下述提供的车险保险业务理赔欺诈识别的实施例描述并不对基于本说明书的其他可扩展到的应用场景构成限制。例如其他的实施场景中，本说明书提供的实施方案同样可以应用到交易纠纷的图像欺诈识别、虚拟货币交易理赔的证据图像等的实施场景中，如购物平台或商家识别买家退货请求上传的物品受损图像是否存在盗图。本领域技术人员可以理解到的是，利用本说明书实施方案应用到其他实施场景中时，相应的“理赔请求”、“理赔图像”等术语的描述可以根据实施场景的不同有相应的描述，如上述示例中的退货请求、物品受损图像可以相应的对应于本说明书实施例方案中的理赔请求、理赔图像。下述的车险保险业务理赔欺诈实施例是一种实施方案的适用场景的示例说明，其他实施场景的实施描述参考本实施例实施处理过程描述，不在进行赘述。具体的一种实施例如图1所示，本说明书提供的一种理赔业务的数据处理方法可以包括：
S0：计算理赔请求对应的理赔图像的理赔特征数据。
用户可以通过终端发起理赔请求。一般的，可以在发起理赔请求时一并上传理赔图像，如通过手机终端拍摄车辆受损图像或者从终端的图库应用中选择图像作为理赔图像。当然，其他具体的实施中也可以先发起理赔请求，在理赔请求受理后再上传理赔图像。本说明书实施例中所述的图像可以包括位图、矢量图等多种格式的图片，也可以包括视频中截取的图像。在一些实施方式中，所述的图像也可以包括视频，视频可以视为 连续图像的集合。
进行理赔业务处理的服务器接收理赔图像后可以实时的进行处理，也可以先进行存储，再基于触发指令进行图像的数据处理。所述的服务器可以采用预设的一种或多种算法来计算理赔图像的理赔特征数据，一般的可以对用户上传的每个理赔图像都计算出其对应的理赔特征数据。特征数据的计算方式有很多，如可以采用一种或多种算法模型抽取图像的特征向量，利用特征向量与历史同类图像的特征向量进行相似程度的对比。具体的例如可以采用尺度不变特征变换(Scale-invariant feature transform，SIFT)算法、加速稳健特征(Speeded Up Robust Features，SURF)、方向梯度直方图(Histogram of Oriented Gradient,HOG)等。当然，也可以采用其他例如利用RGB(红绿蓝三原色)、梯度、灰度等像素点特征，或者其他自定义的算法来提取图像的特征数据。
本说明书提供的一个实施例中，可以采用机器学习的CNN(Convolutional Neural Network，卷积神经管网络)的学习网络抽取图像特征向量。CNN是一种深度的监督学习下的机器学习模型，具有极强的适应性，善于挖掘数据局部特征，提取全局训练特征和分类。本实施例图像特征对比的场景中使用CNN进行图像的特征数据提取，可以更加高效和准确的提取出图像特征，获得更好的识别结果。具体的CNN网络采用的网络模型可以包括CNN网络的变种或改进的机器学习网络模型，其具体实施过程中采用的模型网络层构建可以根据场景需求或设计需求进行相应的设置，在此不做赘述。
需要说明的，在理赔图像的特征数据提取或本说明书其他部分描述的历史同类图像、关联图像等的特征数据提取时，可以使用CNN抽取图像特征向量，也可以使用其他方式，或者CNN与其他方式结合来实现图像特征数据的提取。
S2：计算所述理赔特征数据与历史同类图像的特征数据的相似程度，所述历史同类图像包括与所述理赔请求对应的请求理赔类型相同的理赔类型中所包含的图像信息。
上述描述的理赔图像的特征数据的实施方式，相应的，历史同类图像的特征数据的计算也可以参数上述方式提取获得。一般的，历史同类图像的特征数据的计算与所述理赔图像的特征数据的计算方式可以采用相同的算法，但不排除其他的实施方式中也可以采用不同的特征计算算法，然后经过特征数据的转换/变换/映射等实现理赔图像与历史同类图像的相似程度的计算。
将理赔图像的特征数据(在此可以统一称为理赔特征数据)与历史同类图像的特征数据进行对比计算，确定两者的相似程度。所述的历史同类图像通常的可以包括与所述 理赔请求对应的请求理赔类型相同的理赔类型中所包含的图像信息。如理赔请求对应的理赔类型为车险，则所述历史同类图像可以包括在图库中保险种类同样为车险的历史图像。这些历史图像可以包括以前进行过理赔的案件的图像，也可以包括正在进行理赔处理中的案件的图像，甚至还可以包括从互联网/合作方等获得的图像信息。保险公司可以累积理赔案件的图像数据，存储到图片库中。图片库的图像可以根据险种标记为不同的理赔类型或者相同理赔类型的图像存储到相同理赔类型的存储区块。
所述历史同类图像的特征数据计算可以在进行理赔业务处理时进行图像提取和计算，也可以预先进行计算得到，图像存储到图片库时计算得到特征数据，特征数据一并或单独存储。这样，在进行理赔请求相似程度计算时可以直接调取预先计算好的历史同类图像的特征数据，加快计算速度。
服务器可以基于理赔请求对应的理赔类型，计算相同类型的理赔案件中的图像对应的特征数据，然后将理赔图像的理赔特征数据与历史同类图像的特征数据进行相似度计算。所述的相似度计算具体的可以预先设计相应的计算方式，如特征向量值的比较，差值计算，特征数据关联程度等等。计算得到的相似程度可以使用分值表示，也可以采用其他特征符号表示，如划分相似等级AAA、AA、A、BBB、BB等。
S4：基于得到相似程度确定所述理赔图像中的盗用图像。
计算得到相似程度的表征值，如分值或相似等级后，可以根据相似程度确定某个理赔图像是否为盗用图像。本实施例中可以设置相应的判断表征，如设立分值阈值，当理赔图像与某个历史同类图像的相似程度分值超过阈值时，可以判定该理赔图像为盗用图像。具体的阈值的设定可以根据实际应用场景或处理需求进行设置。
当然，所述的基于相似程度确定理赔图像中是否存在盗用图像时，也可以结合其他数据信息采用设定的规则进行确定，如图像的生成或获取时间信息、图像来源，或者其他图像的标记、属性信息等。或者，其他的实施方式中，也可以采用设置多个阈值，不同阈值对应不同级别的图像盗用/风险级别，如相似度达到90％以上为盗用级别，相似度在60％-76％之间的图像为一级盗用，相似在50％-60％的为二级盗用等。
识别出了盗用图像，则可以进一步的基于盗用图像进行风险识别、监控等处理。
本说明书实施例提供的一种理赔业务的数据处理方法，通过对理赔请求人员上传的理赔图像进行特征提取，将理赔图像与历史同类型理赔的图像进行相似程度判断，判断出理赔图像是否盗用历史已经理赔的案件图像。本说明书实施例可以线下或实时计算申 请理赔图片与同类案件历史图片库相似度，自动、有效、高效的发现盗用历史图像进行理赔的欺诈行为，提高识别处理效率和识别的准确性，降低人工审核成本。
如前所述，识别出了盗用图像，作业人员可以基于具体实际场景中使用盗用图像的情况等来确定所述理赔请求是否存储欺诈行为。这种保险欺诈识别方式对于盗用历史理赔案件图像进行理赔的欺诈行为可以有效的识别，加强了实施方的风险控制，在反欺诈应用中可以有十分显著的识别效果。因此，本说明书还提供所述方法的另一种实施例，如图2所述，具体的还可以包括：
S6：基于所述盗用图像确定所述理赔请求的欺诈识别结果。
具体的如何根据盗用图像确定所述理赔请求是否存在欺诈行为可以进行相应规则的设定。本说明书提供的一个实施方式中，对应保险之类的理赔应用场景，如果识别出用户上传的图片有一个是盗用图像，即可判断该理赔请求存在欺诈行为。因此，所述方法的另一个实施例中，所述基于所述盗用图像确定所述理赔请求的欺诈识别结果可以包括：
S60：若确定至少存在一个盗用图像，则确定所述理赔请求存在欺诈行为。
当然，根据前述描述，也可以采取其他方式确定理赔请求是否存在欺诈行为，如隔绝盗用图像的相似度级别。
所述方法的另一个实施场景中，还可以再结合盗用图像的个数来决策所述理赔请求是否存在欺诈行为。如一些应用场景中采集的理赔图像数量非常多，与以前历史理赔案件的理赔情况(如车险事故中的车辆型号相同、碰撞角度相似、事故现场环境相似等)较为相似，则有可能存在实际理赔图像与历史理赔案件中的图像相似的情况。因此，本实施例中还可以再结合识别出的盗用图像的个数进一步判断是否存在欺诈行为。例如，如果理赔图像与历史理赔案件的图像存在大量相似，超过设定的数量判决阈值，则可以输出理赔请求存在欺诈行为的风险识别结果，可以很好的应用在类似理赔案件中并有效识别出盗用图像的欺诈行为。因此，本说明书提供的所述方法的另一个实施例中，所述基于所述盗用图像确定所述理赔请求的欺诈识别结果包括：
S62：若所述盗用图像的个数达到数量判决阈值，则确定所述理赔请求存在欺诈行为。
一般的，不同类别的理赔案件中所涉及到的图像主体不会或者几乎不会重合或相似，如手机保险和车损保险的图像主体通常是存在明显的不同。但另一些实施场景中，即使 不同类别的理赔案件也可能存在主体相近或相似的情况，如电视屏幕和电脑显示器是两个不同的产品，欺诈行为者有可能使用其中一个产品的破损图片来作为另一个产品进行退货或理赔的理赔图像。因此，本说明书提供的另一个实施例中，可以预先将一些产品或险种的主体相近/相似的理赔类型进行关联，若在上述相同理赔类型中没有发现盗用图像时，则可以进一步的扩大搜索范围，在相关联的理赔类型的图库中进行搜索，查询是否存在盗取其他理赔类型中的图像。这样，可以进一步扩大图像搜索范围，识别利用相近主体的图像进行保险欺诈的盗用图像。具体的，本说明书提供的所述方法的另一个实施例中，如图3所示，若基于所述历史同类图像的特征数据得到相似程度未搜索到盗用图像时，所述方法还包括：
S80：计算所述理赔特征数据与关联图像的特征数据的关联相似程度，所述关联图像包括确定的与所述请求理赔类型相关联的理赔类型中所包含的图像信息；
S82：基于得到所述关联相似程度确定所述理赔图像中的盗用图像。
当然，其他的实施方式中，也可以在即使基于所述历史同类图像的特征数据得到相似程度搜索到了盗用图像时仍然计算所述理赔特征数据与关联图像的特征数据的关联相似程度，确定相关联的理赔类型中的盗用图像。
具体的哪些类型的图像作为相关联的理赔类型，可以根据实际应用场景和设计需求进行设置，可以参考实体产品或服务(险种等)的分类，或自定义的分类，或者一级分类、二级分类等的多级分类进行相应的关联关系设置。
另一个实施场景中，相关联的理赔类型与理赔图像对应的理赔类型不是属于同一个类型，在确定是否为相似度的时候可以相应的调整相似度判决阈值，如可以在相关联的理赔类型中的相似度判决阈值可以低于相同类型中的相似度判决阈值，此时可能确定出多个盗用图像。当然，两者也可以设置相同的取值，或者基于低阶的数据范围采用更严格的判决机制时，相关联的理赔类型中的相似度判决阈值高于相同类型中的相似度判决阈值。在一些实施场景下，不同图像搜索范围确定出的盗用图像可以设置不同的欺诈行为判决机制，可以集合从相关联类型的关联图像中确定的盗用图像的个数来确定所述理赔请求是否存在欺诈行为。具体的，本说明书提供的所述方法的另一个实施例中，如图4所示，所述方法还可以包括：
S10：基于从所述关联图像中确定的盗用图像的个数确定所述理赔请求的欺诈识别结果。
当然，参考前述实施例描述，可以从关联图像中确定出一个盗用图像即确定理赔请求存在欺诈行为，也可以设置其他实施方式，如历史相同图像确定的盗用图像个数和关联图像中确定的盗用图像个数共同决策是否存在欺诈行为。图4中的虚线表示其他的实施例中可以结合其他处理一个或多个处理步骤执行。
本说明书实施例提供的一种理赔业务的数据处理方法，通过对理赔请求人员上传的理赔图像进行特征提取，将理赔图像与历史同类型理赔的图像进行相似程度判断，判断出理赔图像是否盗用历史已经理赔的案件图像。本说明书实施例可以线下或实时计算申请理赔图片与同类案件历史图片库相似度，自动、有效、高效的发现盗用历史图像进行理赔的欺诈行为，提高识别处理效率和识别的准确性，降低人工审核成本。
上述盗用图像的识别或识别是否存在欺诈行为后，可以进一步的进行风险提示或自动进行相应的理赔处理流程。确定没有使用盗用图像，则可以完成后续的理赔流程，如自动理赔或再进行人工的复审。发现盗用历史理赔图像时可以拒绝理赔请求，或者反馈理赔凭证不真实的风险提示或告警灯。
上述所述的方法可以用于服务端一侧的保险欺诈识别，如保险公司或其他服务平台的服务器，也可以应用在其他终端设备中，如PC(personal computer)机、服务器、工控机(工业控制计算机)、移动智能电话、平板电子设备、便携式计算机(例如笔记本电脑等)、个人数字助理(PDA)、或桌面型计算机或智能穿戴设备等。移动通信终端、手持设备、车载设备、可穿戴设备、电视设备、分布式或集成计算设备。如可以应用在保险业务方或服务方或第三方机构的系统服务器中，所述的系统服务器可以包括单独的服务器、服务器集群、分布式系统服务器或者处理设备请求数据的服务器与其他相关联数据处理的系统服务器组合。
本说明书实施例所提供的方法实施例可以在移动终端、计算机终端、服务器或者类似的运算装置中执行。以运行在服务器上为例，图5是应用本发明实施例的一种理赔业务数据处理的服务器的硬件结构框图。如图5所示，服务器10可以包括一个或多个(图中仅示出一个)处理器100(处理器100可以包括但不限于微处理器MCU或可编程逻辑器件FPGA等的处理装置)、用于存储数据的存储器200、以及用于通信功能的传输模块300。本领域普通技术人员可以理解，图5所示的结构仅为示意，其并不对上述电子装置的结构造成限定。例如，服务器10还可包括比图5中所示更多或者更少的组件，例如还可以包括其他的处理硬件，如数据库或多级缓存、GPU(图像处理器)，或者具有与图5所示不同的配置。
存储器200可用于存储应用软件的软件程序以及模块，如本发明实施例中的搜索方法对应的程序指令/模块，处理器100通过运行存储在存储器200内的软件程序以及模块，从而执行各种功能应用以及数据处理，即实现上述导航交互界面内容展示的处理方法。存储器200可包括高速随机存储器，还可包括非易失性存储器，如一个或者多个磁性存储装置、闪存、或者其他非易失性固态存储器。在一些实例中，存储器200可进一步包括相对于处理器100远程设置的存储器，这些远程存储器可以通过网络连接至计算机终端10。上述网络的实例包括但不限于互联网、企业内部网、局域网、移动通信网及其组合。
传输模块300用于经由一个网络接收或者发送数据。上述的网络具体实例可包括计算机终端10的通信供应商提供的无线网络。在一个实例中，传输模块300包括一个网络适配器(Network Interface Controller，NIC)，其可通过基站与其他网络设备相连从而可与互联网进行通讯。在一个实例中，传输模块300可以为射频(Radio Frequency，RF)模块，其用于通过无线方式与互联网进行通讯。
基于上述实施例所述的方法，本说明书还提供一种理赔业务的数据处理装置。所述的装置可以包括使用了本说明书实施例所述方法的系统(包括分布式系统)、软件(应用)、模块、组件、服务器、客户端等并结合必要的实施硬件的设备装置。基于同一创新构思，本说明书提供的一种实施例中的处理装置如下面的实施例所述。由于装置解决问题的实现方案与方法相似，因此本说明书实施例具体的处理装置的实施可以参见前述方法的实施，重复之处不再赘述。尽管以下实施例所描述的装置较佳地以软件来实现，但是硬件，或者软件和硬件的组合的实现也是可能并被构想的。具体的，如图6所示，图6是本说明书提供的一种理赔业务的数据处理装置实施例的模块结构示意图，可以包括：
特征处理模块101，可以用于计算理赔请求对应的理赔图像的理赔特征数据；
相似计算模块102，可以用于计算所述理赔特征数据与历史同类图像的特征数据的相似程度，所述历史同类图像包括与所述理赔请求对应的请求理赔类型相同的理赔类型中所包含的图像信息；
盗用识别模块103，可以用于基于得到相似程度确定所述理赔图像中的盗用图像。
如图7所示，基于前述方法实施例描述，所述装置的另一个实施例中，所述装置还可以包括：
欺诈识别模块104，可以用于基于所述盗用图像确定所述理赔请求的欺诈识别结果。
所述的欺诈识别模块104可以在盗用识别模块103识别出盗用图像时根据设置的规则来决策用户的理赔请求是否存在欺诈行为。
基于前述方法实施例描述，所述装置的另一个实施例中，所述欺诈识别模块104可以包括：
第一识别单元1041，可以用于若确定至少存在一个盗用图像，则确定所述理赔请求存在欺诈行为。
另一个实施例中，所述欺诈识别模块104可以包括：
第二识别单元，可以用于若所述盗用图像的个数达到数量判决阈值，则确定所述理赔请求存在欺诈行为。
如图8所示，所述装置的另一个实施例中，还可以包括：
邻域搜索模块105，可以用于计算所述理赔特征数据与关联图像的特征数据的关联相似程度，所述关联图像包括确定的与所述请求理赔类型相关联的理赔类型中所包含的图像信息；
关联盗用识别模块106，可以用于基于得到所述关联相似程度确定所述理赔图像中的盗用图像。
所述装置的另一个实施例中，所述关联盗用识别模块106可以包括：
个数识别处理单元1061，可以用于基于从所述关联图像中确定的盗用图像的个数确定所述理赔请求的欺诈识别结果。
本说明书实施例提供的服务器或客户端可以在计算机中由处理器执行相应的程序指令来实现，如使用windows操作系统的c++语言在PC端或服务器端实现，或其他例如Linux、android/iOS系统等相对应的应用设计语言集合必要的硬件实现，或者基于量子计算机的处理逻辑实现等。因此，本说明书还提供一种理赔业务的数据处理设备，具体的可以包括处理器以及用于存储处理器可执行指令的存储器，所述处理器执行所述指令时实现：
计算理赔请求对应的理赔图像的理赔特征数据；
计算所述理赔特征数据与历史同类图像的特征数据的相似程度，所述历史同类图像包括与所述理赔请求对应的请求理赔类型相同的理赔类型中所包含的图像信息；
基于得到相似程度确定所述理赔图像中的盗用图像。
当然，所述处理器基于前述实施例描述还可以实现执行其他处理，如基于所述盗用图像确定所述理赔请求的欺诈识别结果等。具体的其他处理设备的实施例描述参照前述相应方法实施例描述，在此不做赘述。
上述的指令可以存储在多种计算机可读存储介质中。所述计算机可读存储介质可以包括用于存储信息的物理装置，可以将信息数字化后再以利用电、磁或者光学等方式的媒体加以存储。本实施例所述的计算机可读存储介质有可以包括：利用电能方式存储信息的装置如，各式存储器，如RAM、ROM等；利用磁能方式存储信息的装置如，硬盘、软盘、磁带、磁芯存储器、磁泡存储器、U盘；利用光学方式存储信息的装置如，CD或DVD。当然，还有其他方式的可读存储介质，例如量子存储器、石墨烯存储器等等。上述所述的装置或服务器或客户端或处理设备中的所涉及的指令同上描述。
上述的处理设备可以具体的为保险服务器或第三方服务机构提供理赔业务反欺诈识别的服务器，所述的服务器可以为单独的服务器、服务器集群、分布式系统服务器或者处理设备请求数据的服务器与其他相关联数据处理的系统服务器组合。因此，本说明书实施例还提供一种具体的服务器产品，所述服务器包括至少一个处理器以及用于存储处理器可执行指令的存储器，所述处理器执行所述指令时实现：
计算理赔请求对应的理赔图像的理赔特征数据；
计算所述理赔特征数据与历史同类图像的特征数据的相似程度，所述历史同类图像包括与所述理赔请求对应的请求理赔类型相同的理赔类型中所包含的图像信息；
基于得到相似程度确定所述理赔图像中的盗用图像。
需要说明的是，本说明书实施例上述所述的装置和处理设备、服务器，根据相关方法实施例的描述还可以包括其他的实施方式。具体的实现方式可以参照方法实施例的描述，在此不作一一赘述。
在本说明书的一个或多个实施例中所述的理赔图像的特征数据计算、历史图像的特征数据计算、盗用图像的确定、是否存在欺诈行为等中的全部或部分的处理，可以采用离线预先构建/处理的方式实现。本说明书不排除其他的实施方式中可以采用在线计算/存储等的实现方式，在计算机能力足够的情况下，可以在线实时进行历史图像的提取、特征计算、图片库累积、欺诈识别等处理。
本说明书中的各个实施例均采用递进的方式描述，各个实施例之间相同相似的 部分互相参见即可，每个实施例重点说明的都是与其他实施例的不同之处。尤其，对于硬件+程序类实施例而言，由于其基本相似于方法实施例，所以描述的比较简单，相关之处参见方法实施例的部分说明即可。
上述对本说明书特定实施例进行了描述。其它实施例在所附权利要求书的范围内。在一些情况下，在权利要求书中记载的动作或步骤可以按照不同于实施例中的顺序来执行并且仍然可以实现期望的结果。另外，在附图中描绘的过程不一定要求示出的特定顺序或者连续顺序才能实现期望的结果。在某些实施方式中，多任务处理和并行处理也是可以的或者可能是有利的。
本说明书实施例提供的一种理赔业务的数据处理方法、装置、设备及服务器，通过对理赔请求人员上传的理赔图像进行特征提取，将理赔图像与历史同类型理赔的图像进行相似程度判断，判断出理赔图像是否盗用历史已经理赔的案件图像。本说明书实施例可以线下或实时计算申请理赔图片与同类案件历史图片库相似度，自动、有效、高效的发现盗用历史图像进行理赔的欺诈行为，提高识别处理效率和识别的准确性，降低人工审核成本。
虽然本申请提供了如实施例或流程图所述的方法操作步骤，但基于常规或者无创造性的劳动可以包括更多或者更少的操作步骤。实施例中列举的步骤顺序仅仅为众多步骤执行顺序中的一种方式，不代表唯一的执行顺序。在实际中的装置或系统服务器产品执行时，可以按照实施例或者附图所示的方法顺序执行或者并行执行(例如并行处理器或者多线程处理的环境)。
尽管本说明书实施例内容中提到图像特征提取、相似度阈值设置、图像或特征数据存在等之类的数据获取、存储、交互、计算、判断等操作和数据描述，但是，本说明书实施例并不局限于必须是符合行业通信标准、标准CNN网络模型或特征提起算法的处理、通信协议和标准数据模型/模板或本说明书实施例所描述的情况。某些行业标准或者使用自定义方式或实施例描述的实施基础上略加修改后的实施方案也可以实现上述实施例相同、等同或相近、或变形后可预料的实施效果。应用这些修改或变形后的数据获取、存储、判断、处理方式等获取的实施例，仍然可以属于本说明书的可选实施方案范围之内。
在20世纪90年代，对于一个技术的改进可以很明显地区分是硬件上的改进(例如，对二极管、晶体管、开关等电路结构的改进)还是软件上的改进(对于方法流程的改进)。然而，随着技术的发展，当今的很多方法流程的改进已经可以视为硬件电路结 构的直接改进。设计人员几乎都通过将改进的方法流程编程到硬件电路中来得到相应的硬件电路结构。因此，不能说一个方法流程的改进就不能用硬件实体模块来实现。例如，可编程逻辑器件(Programmable Logic Device,PLD)(例如现场可编程门阵列(Field Programmable Gate Array，FPGA))就是这样一种集成电路，其逻辑功能由用户对器件编程来确定。由设计人员自行编程来把一个数字系统“集成”在一片PLD上，而不需要请芯片制造厂商来设计和制作专用的集成电路芯片。而且，如今，取代手工地制作集成电路芯片，这种编程也多半改用“逻辑编译器(logic compiler)”软件来实现，它与程序开发撰写时所用的软件编译器相类似，而要编译之前的原始代码也得用特定的编程语言来撰写，此称之为硬件描述语言(Hardware Description Language，HDL)，而HDL也并非仅有一种，而是有许多种，如ABEL(Advanced Boolean Expression Language)、AHDL(Altera Hardware Description Language)、Confluence、CUPL(Cornell University Programming Language)、HDCal、JHDL(Java Hardware Description Language)、Lava、Lola、MyHDL、PALASM、RHDL(Ruby Hardware Description Language)等，目前最普遍使用的是VHDL(Very-High-Speed Integrated Circuit Hardware Description Language)与Verilog。本领域技术人员也应该清楚，只需要将方法流程用上述几种硬件描述语言稍作逻辑编程并编程到集成电路中，就可以很容易得到实现该逻辑方法流程的硬件电路。
控制器可以按任何适当的方式实现，例如，控制器可以采取例如微处理器或处理器以及存储可由该(微)处理器执行的计算机可读程序代码(例如软件或固件)的计算机可读介质、逻辑门、开关、专用集成电路(Application Specific Integrated Circuit，ASIC)、可编程逻辑控制器和嵌入微控制器的形式，控制器的例子包括但不限于以下微控制器：ARC 625D、Atmel AT91SAM、Microchip PIC18F26K20以及Silicone Labs C8051F320，存储器控制器还可以被实现为存储器的控制逻辑的一部分。本领域技术人员也知道，除了以纯计算机可读程序代码方式实现控制器以外，完全可以通过将方法步骤进行逻辑编程来使得控制器以逻辑门、开关、专用集成电路、可编程逻辑控制器和嵌入微控制器等的形式来实现相同功能。因此这种控制器可以被认为是一种硬件部件，而对其内包括的用于实现各种功能的装置也可以视为硬件部件内的结构。或者甚至，可以将用于实现各种功能的装置视为既可以是实现方法的软件模块又可以是硬件部件内的结构。
上述实施例阐明的处理设备、装置、模块或单元，具体可以由计算机芯片或实体实现，或者由具有某种功能的产品来实现。一种典型的实现设备为计算机。具体的， 计算机例如可以为个人计算机、膝上型计算机、车载人机交互设备、蜂窝电话、相机电话、智能电话、个人数字助理、媒体播放器、导航设备、电子邮件设备、游戏控制台、平板计算机、可穿戴设备或者这些设备中的任何设备的组合。
虽然本说明书实施例提供了如实施例或流程图所述的方法操作步骤，但基于常规或者无创造性的手段可以包括更多或者更少的操作步骤。实施例中列举的步骤顺序仅仅为众多步骤执行顺序中的一种方式，不代表唯一的执行顺序。在实际中的装置或终端产品执行时，可以按照实施例或者附图所示的方法顺序执行或者并行执行(例如并行处理器或者多线程处理的环境，甚至为分布式数据处理环境)。术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、产品或者设备不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、产品或者设备所固有的要素。在没有更多限制的情况下，并不排除在包括所述要素的过程、方法、产品或者设备中还存在另外的相同或等同要素。
为了描述的方便，描述以上装置时以功能分为各种模块分别描述。当然，在实施本说明书实施例时可以把各模块的功能在同一个或多个软件和/或硬件中实现，也可以将实现同一功能的模块由多个子模块或子单元的组合实现等。以上所描述的装置实施例仅仅是示意性的，例如，所述单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。另一点，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口，装置或单元的间接耦合或通信连接，可以是电性，机械或其它的形式。
本领域技术人员也知道，除了以纯计算机可读程序代码方式实现控制器以外，完全可以通过将方法步骤进行逻辑编程来使得控制器以逻辑门、开关、专用集成电路、可编程逻辑控制器和嵌入微控制器等的形式来实现相同功能。因此这种控制器可以被认为是一种硬件部件，而对其内部包括的用于实现各种功能的装置也可以视为硬件部件内的结构。或者甚至，可以将用于实现各种功能的装置视为既可以是实现方法的软件模块又可以是硬件部件内的结构。
本发明是参照根据本发明实施例的方法、设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理 设备的处理器以产生一个机器，使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上，使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。
在一个典型的配置中，计算设备包括一个或多个处理器(CPU)、输入/输出接口、网络接口和内存。
内存可能包括计算机可读介质中的非永久性存储器，随机存取存储器(RAM)和/或非易失性内存等形式，如只读存储器(ROM)或闪存(flash RAM)。内存是计算机可读介质的示例。
计算机可读介质包括永久性和非永久性、可移动和非可移动媒体可以由任何方法或技术来实现信息存储。信息可以是计算机可读指令、数据结构、程序的模块或其他数据。计算机的存储介质的例子包括，但不限于相变内存(PRAM)、静态随机存取存储器(SRAM)、动态随机存取存储器(DRAM)、其他类型的随机存取存储器(RAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、快闪记忆体或其他内存技术、只读光盘只读存储器(CD-ROM)、数字多功能光盘(DVD)或其他光学存储、磁盒式磁带，磁带磁磁盘存储或其他磁性存储设备或任何其他非传输介质，可用于存储可以被计算设备访问的信息。按照本文中的界定，计算机可读介质不包括暂存电脑可读媒体(transitory media)，如调制的数据信号和载波。
本领域技术人员应明白，本说明书的实施例可提供为方法、系统或计算机程序产品。因此，本说明书实施例可采用完全硬件实施例、完全软件实施例或结合软件和硬件方面的实施例的形式。而且，本说明书实施例可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储 器等)上实施的计算机程序产品的形式。
本说明书实施例可以在由计算机执行的计算机可执行指令的一般上下文中描述，例如程序模块。一般地，程序模块包括执行特定任务或实现特定抽象数据类型的例程、程序、对象、组件、数据结构等等。也可以在分布式计算环境中实践本说明书实施例，在这些分布式计算环境中，由通过通信网络而被连接的远程处理设备来执行任务。在分布式计算环境中，程序模块可以位于包括存储设备在内的本地和远程计算机存储介质中。
本说明书中的各个实施例均采用递进的方式描述，各个实施例之间相同相似的部分互相参见即可，每个实施例重点说明的都是与其他实施例的不同之处。尤其，对于系统实施例而言，由于其基本相似于方法实施例，所以描述的比较简单，相关之处参见方法实施例的部分说明即可。在本说明书的描述中，参考术语“一个实施例”、“一些实施例”、“示例”、“具体示例”、或“一些示例”等的描述意指结合该实施例或示例描述的具体特征、结构、材料或者特点包含于本说明书实施例的至少一个实施例或示例中。在本说明书中，对上述术语的示意性表述不必须针对的是相同的实施例或示例。而且，描述的具体特征、结构、材料或者特点可以在任一个或多个实施例或示例中以合适的方式结合。此外，在不相互矛盾的情况下，本领域的技术人员可以将本说明书中描述的不同实施例或示例以及不同实施例或示例的特征进行结合和组合。
以上所述仅为本说明书实施例的实施例而已，并不用于限制本说明书实施例。对于本领域技术人员来说，本说明书实施例可以有各种更改和变化。凡在本说明书实施例的精神和原理之内所作的任何修改、等同替换、改进等，均应包含在本说明书实施例的权利要求范围之内。

Claims (14)

  1. 一种理赔业务的数据处理方法，所述方法包括：
    计算理赔请求对应的理赔图像的理赔特征数据；
    计算所述理赔特征数据与历史同类图像的特征数据的相似程度，所述历史同类图像包括与所述理赔请求对应的请求理赔类型相同的理赔类型中所包含的图像信息；
    基于得到相似程度确定所述理赔图像中的盗用图像。
  2. 如权利要求1所述的方法，所述方法还包括：
    基于所述盗用图像确定所述理赔请求的欺诈识别结果。
  3. 如权利要求2所述的方法，所述基于所述盗用图像确定所述理赔请求的欺诈识别结果包括：
    若至少存在一个盗用图像，则确定所述理赔请求存在欺诈行为。
  4. 如权利要求2所述的方法，所述基于所述盗用图像确定所述理赔请求的欺诈识别结果包括：
    若所述盗用图像的个数达到数量判决阈值，则确定所述理赔请求存在欺诈行为。
  5. 如权利要求1所述的方法，若基于所述历史同类图像的特征数据得到相似程度未搜索到盗用图像时，所述方法还包括：
    计算所述理赔特征数据与关联图像的特征数据的关联相似程度，所述关联图像包括确定的与所述请求理赔类型相关联的理赔类型中所包含的图像信息；
    基于得到所述关联相似程度确定所述理赔图像中的盗用图像。
  6. 如权利要求5中所述的方法，所述方法还包括：
    基于从所述关联图像中确定的盗用图像的个数确定所述理赔请求的欺诈识别结果。
  7. 一种理赔业务的数据处理装置，包括：
    特征处理模块，用于计算理赔请求对应的理赔图像的理赔特征数据；
    相似计算模块，用于计算所述理赔特征数据与历史同类图像的特征数据的相似程度，所述历史同类图像包括与所述理赔请求对应的请求理赔类型相同的理赔类型中所包含的图像信息；
    盗用识别模块，用于基于得到相似程度确定所述理赔图像中的盗用图像。
  8. 如权利要求7所述的装置，所述装置还包括：
    欺诈识别模块，用于基于所述盗用图像确定所述理赔请求的欺诈识别结果。
  9. 如权利要求8所述的装置，所述欺诈识别模块包括：
    第一识别模块，用于若确定至少存在一个盗用图像，则确定所述理赔请求存在欺诈 行为。
  10. 如权利要求8所述的装置，所述欺诈识别模块包括：
    第二识别模块，用于若所述盗用图像的个数达到数量判决阈值，则确定所述理赔请求存在欺诈行为。
  11. 如权利要求7所述的装置，所述装置还包括：
    邻域搜索模块，用于计算所述理赔特征数据与关联图像的特征数据的关联相似程度，所述关联图像包括确定的与所述请求理赔类型相关联的理赔类型中所包含的图像信息；
    关联盗用识别模块，用于基于得到所述关联相似程度确定所述理赔图像中的盗用图像。
  12. 如权利要求11所述的装置，所述关联盗用识别模块包括：
    个数识别处理单元，用于基于从所述关联图像中确定的盗用图像的个数确定所述理赔请求的欺诈识别结果。
  13. 一种理赔业务的数据处理设备，包括处理器以及用于存储处理器可执行指令的存储器，所述处理器执行所述指令时实现：
    计算理赔请求对应的理赔图像的理赔特征数据；
    计算所述理赔特征数据与历史同类图像的特征数据的相似程度，所述历史同类图像包括与所述理赔请求对应的请求理赔类型相同的理赔类型中所包含的图像信息；
    基于得到相似程度确定所述理赔图像中的盗用图像。
  14. 一种服务器，包括至少一个处理器以及用于存储处理器可执行指令的存储器，所述处理器执行所述指令时实现：
    计算理赔请求对应的理赔图像的理赔特征数据；
    计算所述理赔特征数据与历史同类图像的特征数据的相似程度，所述历史同类图像包括与所述理赔请求对应的请求理赔类型相同的理赔类型中所包含的图像信息；
    基于得到相似程度确定所述理赔图像中的盗用图像。
PCT/CN2019/075435 2018-05-10 2019-02-19 一种理赔业务的数据处理方法、装置、设备及服务器 WO2019214308A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810440795.6 2018-05-10
CN201810440795.6A CN108573286A (zh) 2018-05-10 2018-05-10 一种理赔业务的数据处理方法、装置、设备及服务器

Publications (1)

Publication Number Publication Date
WO2019214308A1 true WO2019214308A1 (zh) 2019-11-14

Family

ID=63572476

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/075435 WO2019214308A1 (zh) 2018-05-10 2019-02-19 一种理赔业务的数据处理方法、装置、设备及服务器

Country Status (3)

Country Link
CN (1) CN108573286A (zh)
TW (1) TW201947509A (zh)
WO (1) WO2019214308A1 (zh)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111626874A (zh) * 2020-05-25 2020-09-04 泰康保险集团股份有限公司 理赔数据处理方法、装置、设备及存储介质
CN111652581A (zh) * 2020-05-29 2020-09-11 泰康保险集团股份有限公司 理算数据处理方法与装置、存储介质、电子设备
CN112308727A (zh) * 2020-11-30 2021-02-02 泰康保险集团股份有限公司 保险理赔业务处理方法及装置
CN114049227A (zh) * 2021-11-09 2022-02-15 北京优全智汇信息技术有限公司 保险理赔线上结算系统和结算方法
CN116226430A (zh) * 2023-05-08 2023-06-06 深圳市拓保软件有限公司 一种增量式高精度图像过滤方法及系统
CN117196856A (zh) * 2023-08-11 2023-12-08 中国银行保险信息技术管理有限公司 一种理赔信息的处理方法及装置、存储介质、计算机设备

Families Citing this family (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108573286A (zh) * 2018-05-10 2018-09-25 阿里巴巴集团控股有限公司 一种理赔业务的数据处理方法、装置、设备及服务器
CN110009296A (zh) * 2019-02-11 2019-07-12 阿里巴巴集团控股有限公司 理赔业务的处理方法及装置
CN110163761B (zh) * 2019-03-27 2023-05-12 蚂蚁金服（杭州）网络技术有限公司 基于图像处理的可疑项目成员识别方法及装置
CN110135265A (zh) * 2019-04-16 2019-08-16 深圳壹账通智能科技有限公司 案件自动审核方法、装置、计算机设备及存储介质
CN110298454A (zh) * 2019-05-22 2019-10-01 深圳壹账通智能科技有限公司 作业图像的审核方法、装置、计算机设备及存储介质
CN110688514A (zh) * 2019-08-30 2020-01-14 中国人民财产保险股份有限公司 一种保险理赔图像数据的查重方法及装置
CN110689443A (zh) * 2019-09-02 2020-01-14 中国平安财产保险股份有限公司 出险数据处理方法、装置及存储介质、服务器
CN111461905A (zh) * 2020-05-28 2020-07-28 深圳麦亚信科技股份有限公司 车险骗赔规避方法、装置、计算机设备及存储介质
CN112101902A (zh) * 2020-09-15 2020-12-18 京东数字科技控股股份有限公司 数据单处理方法及装置
CN112232971A (zh) * 2020-10-14 2021-01-15 太平金融科技服务（上海）有限公司 反欺诈检测方法、装置、计算机设备和存储介质
CN113763187B (zh) * 2020-12-02 2023-11-07 北京京东振世信息技术有限公司 理赔结果确定方法、理赔请求发送方法和装置
CN114332954A (zh) * 2022-03-10 2022-04-12 太平金融科技服务(上海)有限公司深圳分公司 车险欺诈分析方法、装置、计算机设备和存储介质

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105719188A (zh) * 2016-01-22 2016-06-29 平安科技（深圳）有限公司 基于多张图片一致性实现保险理赔反欺诈的方法及服务器
CN107229947A (zh) * 2017-05-15 2017-10-03 邓昌顺 一种基于动物识别的金融保险方法及系统
CN108573286A (zh) * 2018-05-10 2018-09-25 阿里巴巴集团控股有限公司 一种理赔业务的数据处理方法、装置、设备及服务器

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107679997A (zh) * 2017-10-30 2018-02-09 平安科技（深圳）有限公司 医疗理赔拒付方法、装置、终端设备及存储介质

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105719188A (zh) * 2016-01-22 2016-06-29 平安科技（深圳）有限公司 基于多张图片一致性实现保险理赔反欺诈的方法及服务器
CN107229947A (zh) * 2017-05-15 2017-10-03 邓昌顺 一种基于动物识别的金融保险方法及系统
CN108573286A (zh) * 2018-05-10 2018-09-25 阿里巴巴集团控股有限公司 一种理赔业务的数据处理方法、装置、设备及服务器

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111626874A (zh) * 2020-05-25 2020-09-04 泰康保险集团股份有限公司 理赔数据处理方法、装置、设备及存储介质
CN111626874B (zh) * 2020-05-25 2023-04-25 泰康保险集团股份有限公司 理赔数据处理方法、装置、设备及存储介质
CN111652581A (zh) * 2020-05-29 2020-09-11 泰康保险集团股份有限公司 理算数据处理方法与装置、存储介质、电子设备
CN111652581B (zh) * 2020-05-29 2023-06-13 泰康保险集团股份有限公司 理算数据处理方法与装置、存储介质、电子设备
CN112308727A (zh) * 2020-11-30 2021-02-02 泰康保险集团股份有限公司 保险理赔业务处理方法及装置
CN114049227A (zh) * 2021-11-09 2022-02-15 北京优全智汇信息技术有限公司 保险理赔线上结算系统和结算方法
CN116226430A (zh) * 2023-05-08 2023-06-06 深圳市拓保软件有限公司 一种增量式高精度图像过滤方法及系统
CN116226430B (zh) * 2023-05-08 2023-07-07 深圳市拓保软件有限公司 一种增量式高精度图像过滤方法及系统
CN117196856A (zh) * 2023-08-11 2023-12-08 中国银行保险信息技术管理有限公司 一种理赔信息的处理方法及装置、存储介质、计算机设备

Also Published As

Publication number Publication date
TW201947509A (zh) 2019-12-16
CN108573286A (zh) 2018-09-25

Similar Documents

Publication Publication Date Title
WO2019214308A1 (zh) 一种理赔业务的数据处理方法、装置、设备及服务器
US20200364802A1 (en) Processing method, processing apparatus, user terminal and server for recognition of vehicle damage
WO2020019926A1 (zh) 特征提取模型训练方法、装置、计算机设备及计算机可读存储介质
WO2019214313A1 (zh) 一种车辆定损的交互处理方法、装置、处理设备及客户端
WO2021139324A1 (zh) 图像识别方法、装置、计算机可读存储介质及电子设备
CN108229297B (zh) 人脸识别方法和装置、电子设备、计算机存储介质
CN108229353B (zh) 人体图像的分类方法和装置、电子设备、存储介质、程序
US20170372479A1 (en) Segmentation of objects in videos using color and depth information
US10984225B1 (en) Masked face recognition
WO2015149534A1 (zh) 基于Gabor二值模式的人脸识别方法及装置
WO2018210047A1 (zh) 数据处理方法、数据处理装置、电子设备及存储介质
KR101912748B1 (ko) 확장성을 고려한 특징 기술자 생성 및 특징 기술자를 이용한 정합 장치 및 방법
US9977950B2 (en) Decoy-based matching system for facial recognition
TWI715932B (zh) 車輛損傷識別的處理方法及其處理裝置、車輛定損的資料處理設備、定損處理系統、客戶端及伺服器
WO2021082231A1 (zh) 目标检测方法及装置、电子设备和存储介质
TWI759099B (zh) 一種針對篡改圖像的檢測方法及裝置
CN115578735B (zh) 文本检测方法和文本检测模型的训练方法、装置
WO2020164331A1 (zh) 理赔业务的处理方法及装置
CN114663871A (zh) 图像识别方法、训练方法、装置、系统及存储介质
CN113326773A (zh) 识别模型训练方法、识别方法、装置、设备及存储介质
Jin et al. A deep-learning-based scheme for detecting driver cell-phone use
Deng et al. Attention-aware dual-stream network for multimodal face anti-spoofing
Eid et al. Development of Iris Security System Using Adaptive Quality-Based Template Fusion
US9659234B1 (en) Adaptive selection of scale invariant image feature keypoints
Kar et al. Local centre of mass face for face recognition under varying illumination

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19800291

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19800291

Country of ref document: EP

Kind code of ref document: A1