Image processing method, computing device, storage medium and computer program product

Abstract

The embodiment of the invention provides an image processing method, computing equipment, a storage medium and a computer program product, and the image processing method comprises the steps: determining a plurality of insurance claim settlement images, carrying out the image type detection of each insurance claim settlement image, and determining the image type of each insurance claim settlement image; based on the image type, determining a target insurance claim settlement image from the plurality of insurance claim settlement images, and determining to-be-detected data from the target insurance claim settlement image; based on the to-be-detected data, determining historical detection data corresponding to the to-be-detected data, and performing insurance claim detection on the to-be-detected data based on the historical detection data to obtain an insurance claim detection result, the historical detection data being data passing the insurance claim detection; therefore, an accurate insurance claim settlement detection result is obtained; and the problem of detection errors possibly caused by a manual detection mode is avoided.

Description

Technical Field

[0001] The embodiments of this specification relate to the field of insurance claims technology, and more particularly to an image processing method. One or more embodiments of this specification also relate to a computing device, a computer-readable storage medium, and a computer program product. Background Art

[0002] As insurance services continue to develop, during the insurance claims process, users can submit insurance claim materials for claims, but there may be certain claims risks in this process, such as duplicate reimbursement of invoices, false invoices, etc.

[0003] Currently, the prevention of claims risks is mainly achieved through manual detection, which may result in detection errors. Therefore, how to accurately conduct insurance claims detection has become a technical problem that needs to be solved urgently. Summary of the invention

[0004] In view of this, an embodiment of the present specification provides an image processing method. One or more embodiments of the present specification also relate to an image processing apparatus, a computing device, a computer-readable storage medium and a computer program product to solve the technical defects existing in the prior art.

[0005] According to a first aspect of an embodiment of this specification, there is provided an image processing method, including: Determine a plurality of insurance claim images, and perform image type detection on each insurance claim image to determine the image type of each insurance claim image; Based on the image type, determining a target insurance claim image from the multiple insurance claim images, and determining data to be detected from the target insurance claim image; Based on the data to be detected, historical detection data corresponding to the data to be detected is determined, and insurance claim detection is performed on the data to be detected based on the historical detection data to obtain an insurance claim detection result, wherein the historical detection data is data that has passed the insurance claim detection.

[0006] According to a second aspect of the embodiments of this specification, there is provided an image processing apparatus, including: An image determination module is configured to determine a plurality of insurance claim images, and perform image type detection on each insurance claim image to determine the image type of each insurance claim image; a data determination module configured to determine a target insurance claim image from the plurality of insurance claim images based on the image type, and determine data to be detected from the target insurance claim image; The insurance claim detection module is configured to determine the historical detection data corresponding to the data to be detected based on the data to be detected, and perform insurance claim detection on the data to be detected based on the historical detection data to obtain an insurance claim detection result, wherein the historical detection data is data that has passed the insurance claim detection.

[0007] According to a third aspect of an embodiment of this specification, a computing device is provided, including: Memory and processor; The memory is used to store computer programs / instructions, and the processor is used to execute the computer programs / instructions. When the computer programs / instructions are executed by the processor, the steps of the above-mentioned image processing method are implemented.

[0008] According to a fourth aspect of the embodiments of this specification, a computer-readable storage medium is provided, which stores a computer program / instruction, and when the computer program / instruction is executed by a processor, the steps of the above-mentioned image processing method are implemented.

[0009] According to a fifth aspect of the embodiments of this specification, a computer program product is provided, comprising a computer program / instruction, which implements the steps of the above-mentioned image processing method when executed by a processor.

[0010] One or more embodiments of the present specification provide an image processing method. In the process of insurance claim detection, in order to ensure the accuracy of insurance claim detection, the method can determine a target insurance claim image based on the image type of multiple insurance claim images, and determine the data to be detected from the target insurance claim image; then use the historical detection data that has passed the insurance claim detection to perform insurance claim detection on the data to be detected, so as to detect whether the target insurance claim image has risks such as duplicate claims and false invoices, thereby obtaining accurate insurance claim detection results; avoiding the problem of detection errors that may exist through manual detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 is an application schematic diagram of an image processing method provided by an embodiment of this specification; Figure 2 is a flow chart of an image processing method provided by an embodiment of this specification; Figure 3 is a processing flow chart of an image processing method provided by an embodiment of this specification; Figure 4 is a structural schematic diagram of an image processing device provided by an embodiment of this specification; Figure 5 It is a structural block diagram of a computing device provided by an embodiment of this specification. DETAILED DESCRIPTION

[0012] Many specific details are described in the following description to facilitate a full understanding of this specification. However, this specification can be implemented in many other ways than those described herein, and those skilled in the art can make similar generalizations without violating the connotation of this specification, so this specification is not limited to the specific implementation disclosed below.

[0013] The terms used in one or more embodiments of this specification are only for the purpose of describing specific embodiments, and are not intended to limit one or more embodiments of this specification. The singular forms of "a", "said" and "the" used in one or more embodiments of this specification and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used in one or more embodiments of this specification refers to and includes any or all possible combinations of one or more associated listed items.

[0014] It should be understood that although the terms first, second, etc. may be used to describe various information in one or more embodiments of this specification, this information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of one or more embodiments of this specification, the first may also be referred to as the second, and similarly, the second may also be referred to as the first. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0015] In addition, it should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in one or more embodiments of this specification are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0016] First, the terms involved in one or more embodiments of this specification are explained.

[0017] OCR (Optical Character Recognition): refers to optical character recognition, which is a text recognition technology that can convert text in an image into machine-encoded text.

[0018] CNN (Convolutional Neural Networks): refers to convolutional neural networks.

[0019] PaddleOCR: A high-performance optical character recognition tool developed based on a deep learning framework.

[0020] A regular expression is a pattern used to match character combinations in a string. It uses specific grammatical rules to describe the structure of a string, thereby achieving efficient string search, matching, and replacement operations. Regular expressions are widely used in text processing and can be used to verify data, find and replace text, and extract information.

[0021] SQL (Structured Query Language): refers to structured query language.

[0022] NoSQL: A JSON-like query language for operating its database, which allows complex queries to filter, sort, and limit results.

[0023] GraphQL: A data query and manipulation language for APIs.

[0024] With the continuous development of insurance services, during the insurance claim process, users can submit insurance claim materials for claims, but there may be certain claims risks in this process. During the insurance claim process, the submission of duplicate invoices and false invoices is a long-standing problem; insurance companies need to verify the authenticity and uniqueness of each invoice to prevent duplicate claims and fraud. However, the current manual review method is inefficient and prone to omissions, especially when processing a large number of claims applications, the accuracy and speed of manual review are difficult to guarantee; this not only increases the operating costs of insurance companies, but may also lead to claims errors and damage the interests of the company. Specifically, the disadvantages of the manual review method are: first, it is time-consuming; manual classification and identification are time-consuming, and it is impossible to process a large number of report pictures on a large scale; second, it is error-prone, and manual review is prone to misclassification and misidentification.

[0025] In response to the above problems, this specification provides an OCR-based system, which performs risk detection through OCR technology; however, the disadvantages of this system are: first, the classification accuracy is not high, the classification accuracy of simple image processing technology is not high, and it is easy to misclassify; second, the recognition accuracy is not high, and the recognition accuracy of basic OCR technology is not high, especially in low-quality images.

[0026] Based on this, in this specification, an image processing method is provided. This specification also involves one or more embodiments of this specification and also involves an image processing device, a computing device, a computer-readable storage medium and a computer program product, which are described in detail one by one in the following embodiments.

[0027] See also Figure 1 , Figure 1 A schematic diagram of an application of an image processing method provided according to an embodiment of the present specification is shown. Figure 1 It can be seen that the image processing method provided in the embodiment of the present application can be applied to Figure 1 The application scenarios shown are not limited to these. Figure 1 In the application scenario shown, the method is applied to a server 10, and the server 10 can be connected to one or more client devices 20 through a local area network connection, a wide area network connection, an Internet connection, or other types of data networks. The client devices 20 here may include but are not limited to: smart phones, tablet computers, laptops, PDAs, personal computers, smart home devices, vehicle-mounted devices, etc. The client device 20 can interact with the user through a graphical user interface to implement the method provided in the embodiments of this specification.

[0028] In the embodiment of this specification, the system composed of the client device 20 and the server 10 can perform the following steps: the client device 20 executes to send the claim picture uploaded by the user through the graphical user interface to the server 10; the server 10 performs operations such as picture classification, invoice number extraction, and invoice database query; wherein, picture classification refers to: using the pre-trained CNN model to classify the picture and output the classification result of each picture; invoice number extraction refers to: using OCR technology to recognize the text in the picture, and using regular expressions to extract the invoice number from the recognized text; invoice database query refers to: constructing an SQL query statement based on the invoice number extracted by the regular expression, and querying whether the extracted invoice number exists in the database based on the SQL query statement, and determining whether there is a risk of duplicate claims by whether the invoice number exists in the database. This effectively prevents the risk of duplicate invoices in the claims process and improves the accuracy and efficiency of claims.

[0029] See also Figure 2 , Figure 2 A flowchart of an image processing method provided according to an embodiment of the present specification is shown, which specifically includes the following steps.

[0030] Step 202: Determine a plurality of insurance claim images, and perform image type detection on each insurance claim image to determine the image type of each insurance claim image.

[0031] The insurance claim image may be understood as image data required in the insurance claim process; for example, the insurance claim image may be an invoice image, a contract image, a medical record image, an injury assessment certificate image, etc. The multiple insurance claim images may be insurance claim images of multiple image types, and each image type of insurance claim image may include one or more insurance claim images.

[0032] The image type can be understood as the type corresponding to the insurance claim image. For example, the image type can be an invoice type, a contract type, a medical record type, an injury assessment type, etc., without specific limitations here; each type of insurance claim image can include one or more, for example, an insurance claim image of the invoice type can be one or more invoice images.

[0033] In one or more embodiments provided in this specification, in claims risk control, duplicate invoice risk or false invoice risk is a common problem that may cause economic losses to the insurance company, so risk detection is required. In the process of risk detection, it is necessary to identify and classify report images; however, due to the wide variety of report images, how to quickly and accurately identify and classify these images becomes a problem that needs to be solved; based on this, in the process of performing image type detection on each insurance claim image, the method can use an image classification model to achieve fast and accurate classification, and the specific implementation method is as follows.

[0034] The performing image type detection on each insurance claim image to determine the image type of each insurance claim image includes: The insurance claim images are input into an image classification model, and the image classification model is used to perform image type detection on the insurance claim images to obtain the image type of the insurance claim images.

[0035] Among them, the image classification model can be understood as a model used to classify insurance claim images. The image classification model can be a CNN model, a deep learning model, etc., and no specific restrictions are made here.

[0036] Taking the application of the image processing method provided in this specification in the claims risk control scenario as an example, the image processing method is described; wherein the image classification model can be a CNN model. Based on this, this method can be classified based on the CNN model, specifically in the following manner: input the pictures provided by the user (i.e., insurance claims images) into the pre-trained CNN model, classify the pictures using the pre-trained CNN model, and output the classification results of each picture.

[0037] It should be noted that, in the process of detecting the image type of each insurance claim image, this method can use image classification tools to achieve fast and accurate classification, and the specific implementation method is as follows.

[0038] The performing image type detection on each insurance claim image to determine the image type of each insurance claim image includes: The image classification tool is used to perform image type detection on each of the insurance claim images to obtain the image type of each of the insurance claim images.

[0039] The image classification tool can be understood as an application or algorithm for realizing image type detection, and no specific limitation is made here. Based on this, the method can call the image classification tool to classify the pictures (i.e., insurance claim pictures) provided by the user to obtain the classification results of each picture.

[0040] In one or more embodiments provided in this specification, the image processing method provided in this specification can be applied to the server and interact with the client to achieve insurance claim risk detection; the specific implementation method is as follows: The determining of a plurality of insurance claim images comprises: Receiving the multiple insurance claim images sent by the client, wherein the multiple insurance claim images are sent by the client when the user performs an insurance claim operation on the insurance claim interface; Among them, the insurance claim interface can be understood as a human-computer interaction interface used for making insurance claims. For example, the insurance claim interface can be an application program interface or a web page.

[0041] Continuing with the above example, the user uploads the pictures needed for insurance claims through the client's mobile application or web page. The client sends the pictures to the system (i.e., the server), which can receive and store these pictures; thereby realizing the collection of user claim pictures and facilitating subsequent insurance claim risk detection.

[0042] In one or more embodiments provided in this specification, in the claims risk control, in order to ensure image format statistics and improve the efficiency of subsequent image processing, the insurance claims images to be processed uploaded by the user can be pre-processed to obtain multiple insurance claims images. The specific implementation method is as follows: A plurality of insurance claim images to be processed are determined, and each of the insurance claim images to be processed is preprocessed to obtain the plurality of insurance claim images, wherein the preprocessing includes image scaling, image normalization, image rotation and other processing.

[0043] Continuing with the above example, this method can load images, read each image from the stored images, and preprocess each image (such as scaling and normalization) to obtain the processed images; the processed images can be input into the CNN model for classification processing.

[0044] Step 204: Based on the image type, determine a target insurance claim image from the multiple insurance claim images, and determine the data to be detected from the target insurance claim image.

[0045] Among them, the target insurance claim image can be understood as an image among multiple insurance claim images that needs to be detected for insurance claim risk; when the present method is applied to a duplicate invoice or false invoice detection scenario, the target insurance claim image can be an invoice image; that is, in this specification, the target insurance claim image can be an insurance claim invoice image; when the present method is applied to a false evidence detection scenario, the target insurance claim image can be a medical record image or an injury assessment certificate image, etc.

[0046] The data to be detected is understood to be the data that needs to be detected for insurance claim risk among the multiple data contained in the target insurance claim image; since the target insurance claim image may include a large amount of data (such as text data), in order to improve the efficiency of insurance claim risk detection, the method can select data associated with insurance claim risk from multiple data as the data to be detected; for example, when the method is applied to the duplicate invoice or false invoice detection scenario, the data to be detected can be the invoice number contained in the invoice image. In other words, the data to be detected can be the invoice number. For example, when the method is applied to the false evidence detection scenario, the data to be detected can be the QR code or barcode contained in the medical record image.

[0047] In one or more embodiments provided in this specification, in order to improve the efficiency of insurance claim risk detection, it is necessary to quickly determine a target insurance claim image from multiple insurance claim images, and the specific implementation method is as follows.

[0048] The step of determining a target insurance claim image from the plurality of insurance claim images based on the image type includes: Determine, among the multiple insurance claim images, an insurance claim image whose image type is consistent with the target image type as a target insurance claim image; or The risk level of each insurance claim image is determined based on the image type, and an insurance claim image whose risk level is consistent with a target risk level is determined as a target insurance claim image.

[0049] Among them, the target image type can be an image type associated with the insurance claim risk. For example, when the present method is applied to the false invoice or duplicate invoice detection scenario, the target image type can be an invoice type. Based on this, among multiple insurance claim images, an insurance claim image whose image type is an invoice type can be determined as a target insurance claim image.

[0050] The risk level may be information used to characterize the degree of risk corresponding to the image type; the risk level may be high risk, low risk, no risk, etc., and is not specifically limited here.

[0051] The target risk level may be understood as a pre-set risk level for screening insurance claim images that require insurance claim risk detection. For example, the target risk level may be high risk, or the target risk level may be low risk and high risk.

[0052] The image processing method provided in this specification can assign a corresponding risk level to each image type. The higher the degree of association with the insurance claim risk of the image type, the higher its risk level. For example, when this method is applied to the false invoice or duplicate invoice detection scenario, the invoice type has a high degree of association with the risks of false invoice or duplicate invoice, so the invoice type can correspond to a high risk; while the contract type and medical record type have a low degree of association with the risks of false invoice or duplicate invoice, so the contract type and medical record type can correspond to a low risk. Based on this, this method can determine the risk level of each insurance claim image based on the image type, and determine the insurance claim image whose risk level is consistent with the target risk level (e.g., high risk) as the target insurance claim image.

[0053] In one or more embodiments provided in this specification, in claims risk control, duplicate invoice risk or false invoice risk is a common problem that may cause economic losses to the insurance company, and therefore risk detection is required. In the process of risk detection, accurate recognition of the invoice number is required. In order to accurately extract the invoice number from the image, accurate extraction of the invoice information is required. OCR technology can be used to recognize text from the image. In turn, how to accurately extract key information such as the invoice number, invoice code, and hospital name from the OCR-recognized text becomes a problem that needs to be solved. In the process of determining the data to be detected from the target insurance claim image in this method, regular expressions can be used to accurately determine the data to be detected from the target insurance claim image. The specific implementation method is as follows.

[0054] The determining the data to be detected from the target insurance claim image includes: Performing text recognition on the target insurance claim image using a text recognition algorithm to obtain a plurality of image texts contained in the target insurance claim image; The data to be detected is identified from the plurality of image texts using a regular expression for the image text, wherein the data to be detected is an image text whose data format is consistent with the regular expression.

[0055] Among them, the text recognition algorithm can be understood as an algorithm used to extract text data from an image. For example, the text recognition algorithm can be OCR technology, a deep learning model, etc.

[0056] Continuing with the above example, the text recognition algorithm is Paddle OCR technology; based on this, in the process of invoice number recognition, this method first performs Paddle OCR recognition: loads the image recognized as the invoice type (that is, the target insurance claim image), and uses Paddle OCR technology to recognize the text in the image (that is, multiple image texts).

[0057] Then, perform Extract Invoice Number: Use regular expressions to extract the invoice number from the recognized text.

[0058] In one or more embodiments provided in this specification, in the process of information extraction, a regular expression extraction method can be used: by defining a regular expression matching pattern, the invoice number and associated text (such as invoice code, hospital name, etc.) are extracted from the text recognized by OCR; this facilitates subsequent text processing operations based on the associated text (such as generating prompt information based on the associated text, storing the associated text in a database, etc.).

[0059] Step 206: Based on the data to be detected, determine the historical detection data corresponding to the data to be detected, and perform insurance claim detection on the data to be detected based on the historical detection data to obtain an insurance claim detection result, wherein the historical detection data is data that has passed the insurance claim detection.

[0060] The historical detection data may be understood as data to be detected for historical insurance claims detection; for example, historical invoice numbers for which insurance claims detection has been performed historically.

[0061] Insurance claim detection can be understood as insurance claim risk detection for insurance claim services. For example, the insurance claim detection can be duplicate invoice detection, false invoice detection, false evidence detection, false claim material detection, etc.

[0062] The insurance claim detection result can be understood as the detection result of whether the target insurance claim image has an insurance claim risk, which is determined after the insurance claim detection; for example, when the present method is applied to a false invoice or duplicate invoice detection scenario, the insurance claim detection result can be the detection result of whether it is a false invoice or a duplicate invoice; for example, when the present method is applied to a false evidence detection scenario, the insurance claim detection result can be the detection result of whether it is false evidence.

[0063] In one or more embodiments provided in this specification, during the process of insurance claim detection, it is necessary to obtain historical invoice numbers from the database, so it is necessary to query and update the invoice database: how to effectively query the invoice database and prevent duplicate claims becomes a problem that needs to be solved. Based on this, this method uses a data query statement to efficiently query the historical detection data corresponding to the data to be detected from the data storage unit; the specific implementation method is as follows.

[0064] The determining, based on the data to be detected, the historical detection data corresponding to the data to be detected includes: Generate a data query statement using the data information of the data to be detected; By executing the data query statement, the historical detection data corresponding to the data to be detected is queried from the data storage unit.

[0065] Among them, the data query statement can be understood as a statement used to query data from a data storage unit. For example, the data query statement can be an SQL query statement, a NoSQL statement, a GraphQL statement, etc.

[0066] The data storage unit can be understood as a unit for storing historical detection data. The data storage unit can be a database, a server, a local disk, etc., and no specific limitation is made here; for example, the data storage unit can be an invoice database for storing invoice numbers.

[0067] Continuing with the above example, in the process of detecting fake or duplicate invoices, it is necessary to query the invoice database. In the process of querying the database, it is necessary to connect to the invoice database and build an SQL query statement based on the invoice number extracted by the regular expression. Use the SQL query statement to query the historical invoice numbers stored in the database.

[0068] In one or more embodiments provided in this specification, by determining whether the data storage unit stores historical detection data that is consistent with the data to be detected, it is efficiently determined whether there is an insurance claim risk; the specific implementation process is as follows.

[0069] The performing insurance claim detection on the data to be detected based on the historical detection data to obtain the insurance claim detection result includes: In the case of determining that the data storage unit stores the historical detection data, determining that the multiple insurance claim images have insurance claim risks, and taking the existence of insurance claim risks as the insurance claim detection result, wherein the historical detection data is consistent with the data to be detected; When it is determined that the data storage unit does not store the historical detection data, it is determined that the multiple insurance claim images do not have insurance claim risks, the data to be detected is stored as historical detection data in the data storage unit, and the absence of insurance claim risks is used as the insurance claim detection result.

[0070] Continuing with the above example, in the process of detecting fake or duplicate invoices, it is necessary to query the invoice database. In the process of querying the database, it is necessary to connect to the invoice database, and build an SQL query statement based on the invoice number extracted by the regular expression, and then query whether the extracted invoice number exists in the database based on the SQL query statement.

[0071] Determine whether it exists: If the invoice number exists in the database, it is determined that there is a risk of duplicate claims; if it does not exist, the invoice information is added to the database.

[0072] In one or more embodiments provided in this specification, after performing insurance claim detection on the to-be-detected data based on the historical detection data and obtaining the insurance claim detection result, the method includes: Insurance claim prompt information is generated based on the insurance claim detection result, and the insurance claim detection result is sent to the client.

[0073] Among them, the insurance claim prompt information can be understood as information used to indicate the insurance claim detection results. For example, the insurance claim prompt information can be a text message, email, instant messaging software message, etc., and no specific restrictions are made here.

[0074] Continuing with the above example, when the insurance claim detection result is obtained, if the invoice number exists in the database, it will indicate the risk of duplicate claims. By generating duplicate claim risk warning information (i.e., insurance claim warning information) and sending it to the client, the claims personnel will be reminded to pay attention to the risk of duplicate claims.

[0075] Alternatively, when obtaining the insurance claim detection result, if the invoice number does not exist in the database, it is prompted that there is no risk of duplicate claims. By generating a prompt message (i.e., the insurance claim prompt message) and sending it to the client, the claims personnel are reminded that there is no risk of duplicate claims, and the invoice information is added to the database, thereby updating the database.

[0076] One or more embodiments of the present specification provide an image processing method. In the process of insurance claim detection, in order to ensure the accuracy of insurance claim detection, the method can determine a target insurance claim image based on the image type of multiple insurance claim images, and determine the data to be detected from the target insurance claim image; then use the historical detection data that has passed the insurance claim detection to perform insurance claim detection on the data to be detected, so as to detect whether the target insurance claim image has risks such as duplicate claims and false invoices, thereby obtaining accurate insurance claim detection results; avoiding the problem of detection errors that may exist through manual detection.

[0077] The following combination Figure 3 , taking the application of the image processing method provided in this specification in the insurance claim risk control scenario as an example, the image processing method is further explained. Figure 3 A processing flow chart of an image processing method provided by an embodiment of the present specification is shown, which specifically includes the following steps.

[0078] Step 302: Collect user claim pictures.

[0079] First, the user uploads the claim picture through the client's mobile application interface or web page.

[0080] Secondly, the server receives and stores these images.

[0081] Step 304: The picture classification model performs picture classification.

[0082] First, load the images: read each image from the stored images and perform preprocessing (such as scaling and normalization).

[0083] Secondly, use the CNN model for classification: use the pre-trained CNN model to classify the pictures and output the classification results of each picture to determine the picture type of each picture.

[0084] The image types include but are not limited to: invoice type, contract type, medical record type, etc.

[0085] Step 306: Paddle OCR performs data recognition.

[0086] Specifically, perform Paddle OCR recognition: load an image identified as an invoice type, and use Paddle OCR technology to recognize the text in the image.

[0087] The picture of this invoice type may be a normal photo of the electronic invoice.

[0088] Step 308: Extract the invoice number.

[0089] First, perform regular expression extraction: by defining regular expression matching patterns, extract invoice number, invoice code, hospital name and other information from the OCR-recognized text.

[0090] Secondly, the extracted text information is stored for subsequent use.

[0091] Step 310: Query the invoice database.

[0092] In the process of detecting false or duplicate invoices, it is necessary to query the invoice database. In the process of querying the database, it is necessary to connect to the invoice database, build an SQL query statement based on the invoice number extracted by the regular expression, and query whether the extracted invoice number exists in the database based on the SQL query statement.

[0093] Step 312: Determine whether it is a hit, if so, execute step 314, if not, execute step 316.

[0094] Step 314: Remind of the risk of the claims operation.

[0095] Specifically, if the invoice number exists in the database, it indicates that there is a risk of duplicate claims, reminding the claims adjuster to pay attention to the risk of duplicate claims.

[0096] Step 316: Add to the invoice database.

[0097] Specifically, if the invoice number exists in the database, the invoice information is added to the database, thereby updating the database to facilitate subsequent risk detection based on the updated invoice number in the database.

[0098] Based on the above steps, it can be seen that the image processing method provided in this manual provides a method for risk insight and prevention of duplicate invoices in claims risk control. This method can effectively prevent the risk of duplicate invoices in the claims process and improve the accuracy and efficiency of claims. In addition, this method integrates the above key technologies into a complete process, realizing the full process automation from image collection, classification, recognition, information extraction to database query and reminder.

[0099] Corresponding to the above method embodiment, this specification also provides an image processing device embodiment, Figure 4 FIG. 2 shows a schematic diagram of the structure of an image processing device provided by an embodiment of the present specification. Figure 4 As shown, the device comprises: The image determination module 402 is configured to determine a plurality of insurance claim images, and perform image type detection on each insurance claim image to determine the image type of each insurance claim image; The data determination module 404 is configured to determine a target insurance claim image from the plurality of insurance claim images based on the image type, and determine the data to be detected from the target insurance claim image; The insurance claim detection module 406 is configured to determine the historical detection data corresponding to the data to be detected based on the data to be detected, and perform an insurance claim detection on the data to be detected based on the historical detection data to obtain an insurance claim detection result, wherein the historical detection data is data that has passed the insurance claim detection.

[0100] Optionally, the image determination module 402 is further configured to: The insurance claim images are input into an image classification model, and the image classification model is used to perform image type detection on the insurance claim images to obtain the image type of the insurance claim images.

[0101] Optionally, the data determination module 404 is further configured to: Determine, among the multiple insurance claim images, an insurance claim image whose image type is consistent with the target image type as a target insurance claim image; or The risk level of each insurance claim image is determined based on the image type, and an insurance claim image whose risk level is consistent with a target risk level is determined as a target insurance claim image.

[0102] Optionally, the data determination module 404 is further configured to: Performing text recognition on the target insurance claim image using a text recognition algorithm to obtain a plurality of image texts contained in the target insurance claim image; The data to be detected is identified from the multiple image texts by using a regular expression for the image text, wherein the data to be detected is an image text whose data format is consistent with the regular expression.

[0103] Optionally, the insurance claim detection module 406 is further configured to: Generate a data query statement using the data information of the data to be detected; By executing the data query statement, the historical detection data corresponding to the data to be detected is queried from the data storage unit.

[0104] Optionally, the image determination module 402 is further configured to: Receiving the multiple insurance claim images sent by the client, wherein the multiple insurance claim images are sent by the client when the user performs an insurance claim operation on the insurance claim interface; Optionally, the image processing device further includes a result sending module configured to: Insurance claim prompt information is generated based on the insurance claim detection result, and the insurance claim detection result is sent to the client.

[0105] Optionally, the insurance claim detection module 406 is further configured to: In the case of determining that the data storage unit stores the historical detection data, determining that the multiple insurance claim images have insurance claim risks, and taking the existence of insurance claim risks as the insurance claim detection result, wherein the historical detection data is consistent with the data to be detected; When it is determined that the data storage unit does not store the historical detection data, it is determined that the multiple insurance claim images do not have insurance claim risks, the data to be detected is stored as historical detection data in the data storage unit, and the absence of insurance claim risks is used as the insurance claim detection result.

[0106] Optionally, the target insurance claim image is an insurance claim invoice image, and the data to be detected is an invoice number.

[0107] One or more embodiments of the present specification provide an image processing device. In the process of insurance claim detection, in order to ensure the accuracy of insurance claim detection, the method can determine a target insurance claim image based on the image type of multiple insurance claim images, and determine the data to be detected from the target insurance claim image; then use the historical detection data that has passed the insurance claim detection to perform insurance claim detection on the data to be detected, so as to detect whether the target insurance claim image has risks such as repeated claims and false invoices, thereby obtaining accurate insurance claim detection results; avoiding the problem of detection errors that may exist through manual detection.

[0108] The above is a schematic scheme of an image processing device of this embodiment. It should be noted that the technical scheme of the image processing device and the technical scheme of the above-mentioned image processing method belong to the same concept, and the details not described in detail in the technical scheme of the image processing device can be referred to the description of the technical scheme of the above-mentioned image processing method.

[0109] Figure 5 The structure block diagram of a computing device 500 provided according to an embodiment of the present specification is shown. The components of the computing device 500 include but are not limited to a memory 510 and a processor 520. The processor 520 is connected to the memory 510 via a bus 530, and the database 550 is used to store data.

[0110] The computing device 500 also includes an access device 540 that enables the computing device 500 to communicate via one or more networks 560. Examples of these networks include a public switched telephone network (PSTN), a local area network (LAN), a wide area network (WAN), a personal area network (PAN), or a combination of communication networks such as the Internet. The access device 540 may include one or more of any type of network interface (e.g., a network interface card (NIC)) that is wired or wireless, such as an IEEE 802.11 wireless local area network (WLAN) wireless interface, a world-wide interoperability for microwave access (Wi-MAX) interface, an Ethernet interface, a universal serial bus (USB) interface, a cellular network interface, a Bluetooth interface, and a near field communication (NFC).

[0111] In one embodiment of the present specification, the above components of the computing device 500 and Figure 5 Other components not shown in the figure may also be connected to each other, for example, via a bus. It should be understood that Figure 5 The computing device structure block diagram shown is only for the purpose of illustration, and is not intended to limit the scope of this specification. Those skilled in the art can add or replace other components as needed.

[0112] The computing device 500 may be any type of stationary or mobile computing device, including a mobile computer or mobile computing device (e.g., a tablet computer, a personal digital assistant, a laptop computer, a notebook computer, a netbook, etc.), a mobile phone (e.g., a smart phone), a wearable computing device (e.g., a smart watch, smart glasses, etc.), or other types of mobile devices, or a stationary computing device such as a desktop computer or a personal computer (PC). The computing device 500 may also be a mobile or stationary server.

[0113] The processor 520 is used to execute the following computer executable instructions, which implement the steps of the above-mentioned image processing method when executed by the processor.

[0114] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the computing device embodiment, since it is basically similar to the image processing method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the image processing method embodiment.

[0115] An embodiment of the present specification further provides a computer-readable storage medium storing a computer program / instruction, which implements the steps of the above-mentioned image processing method when executed by a processor.

[0116] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the computer-readable storage medium embodiment, since it is basically similar to the image processing method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the image processing method embodiment.

[0117] An embodiment of the present specification further provides a computer program product, including a computer program / instruction, which implements the steps of the above-mentioned image processing method when executed by a processor.

[0118] The above is a schematic solution of a computer program product of this embodiment. It should be noted that the technical solution of the computer program product and the technical solution of the above-mentioned image processing method belong to the same concept, and the details not described in detail in the technical solution of the computer program product can be referred to the description of the technical solution of the above-mentioned image processing method.

[0119] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0120] The computer instructions include computer program codes, which may be in source code form, object code form, executable files or some intermediate forms, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium may be appropriately increased or decreased according to the requirements of patent practice. For example, in some regions, according to patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0121] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity of description, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the embodiments of this specification are not limited by the order of the actions described, because according to the embodiments of this specification, some steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the embodiments of this specification.

[0122] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0123] The preferred embodiments of this specification disclosed above are only used to help explain this specification. The optional embodiments do not describe all the details in detail, nor do they limit the invention to only the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of the embodiments of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the embodiments of this specification, so that technicians in the relevant technical field can well understand and use this specification. This specification is only limited by the claims and their full scope and equivalents.

Claims

1. An image processing method, characterized in that: include: Determine a plurality of insurance claim images, and perform image type detection on each insurance claim image to determine the image type of each insurance claim image; Based on the image type, determining a target insurance claim image from the multiple insurance claim images, and determining data to be detected from the target insurance claim image; Based on the data to be detected, historical detection data corresponding to the data to be detected is determined, and insurance claim detection is performed on the data to be detected based on the historical detection data to obtain an insurance claim detection result, wherein the historical detection data is data that has passed the insurance claim detection.

2. The image processing method according to claim 1, characterized in that: The performing image type detection on each insurance claim image to determine the image type of each insurance claim image includes: The insurance claim images are input into an image classification model, and the image classification model is used to perform image type detection on the insurance claim images to obtain the image type of the insurance claim images.

3. The image processing method according to claim 1, characterized in that: The step of determining a target insurance claim image from the plurality of insurance claim images based on the image type includes: Determine, among the multiple insurance claim images, an insurance claim image whose image type is consistent with the target image type as a target insurance claim image; or The risk level of each insurance claim image is determined based on the image type, and an insurance claim image whose risk level is consistent with a target risk level is determined as a target insurance claim image.

4. The image processing method according to claim 1, characterized in that: The determining the data to be detected from the target insurance claim image includes: Performing text recognition on the target insurance claim image using a text recognition algorithm to obtain a plurality of image texts contained in the target insurance claim image; The data to be detected is identified from the plurality of image texts by using a regular expression for the image texts.

5. The image processing method according to claim 1, characterized in that: The determining, based on the data to be detected, the historical detection data corresponding to the data to be detected includes: Generate a data query statement using the data information of the data to be detected; By executing the data query statement, the historical detection data corresponding to the data to be detected is queried from the data storage unit.

6. The image processing method according to claim 1, characterized in that: The determining of a plurality of insurance claim images comprises: Receiving the multiple insurance claim images sent by the client, wherein the multiple insurance claim images are sent by the client when the user performs an insurance claim operation on the insurance claim interface; After performing insurance claim detection on the data to be detected based on the historical detection data and obtaining the insurance claim detection result, the method includes: Insurance claim prompt information is generated based on the insurance claim detection result, and the insurance claim detection result is sent to the client.

7. The image processing method according to any one of claims 1 or 5, characterized in that: The performing insurance claim detection on the data to be detected based on the historical detection data to obtain the insurance claim detection result includes: In the case of determining that the data storage unit stores the historical detection data, determining that the multiple insurance claim images have insurance claim risks, and taking the existence of insurance claim risks as the insurance claim detection result, wherein the historical detection data is consistent with the data to be detected; When it is determined that the data storage unit does not store the historical detection data, it is determined that the multiple insurance claim images do not have insurance claim risks, the data to be detected is stored as historical detection data in the data storage unit, and the absence of insurance claim risks is used as the insurance claim detection result.

8. The image processing method according to any one of claims 1 to 6, characterized in that: The target insurance claim image is an insurance claim invoice image, and the data to be detected is an invoice number.

9. A computing device, characterized in that include: Memory and processor; The memory is used to store computer programs / instructions, and the processor is used to execute the computer programs / instructions. When the computer program / instructions are executed by the processor, the steps of the method according to any one of claims 1 to 8 are implemented.

10. A computer-readable storage medium, characterized in that: It stores a computer program / instruction, which implements the steps of the method described in any one of claims 1 to 8 when executed by a processor.

11. A computer program product, characterized in that The method comprises a computer program / instruction which, when executed by a processor, implements the steps of the method according to any one of claims 1 to 8.

Images