Image Processing Method, Apparatus, Computer Device, and Storage Medium

Abstract

An embodiment of the present application discloses an image processing method, apparatus, computer device, and storage medium. The image processing method includes: displaying a transaction monitoring interface, where the transaction monitoring interface includes historical transaction data and an image display control; when the image display control is selected, displaying an abnormal result interface; the abnormal result interface includes one or more abnormal dimensions, and each abnormal dimension includes one or more abnormal transaction data images. The abnormal transaction data images under each abnormal dimension are obtained by detecting the historical transaction data using a detection rule corresponding to each abnormal dimension; when there is a preview operation for a target abnormal dimension, displaying a preview interface for the target abnormal dimension, and displaying one or more abnormal transaction data images under the target abnormal dimension in the preview interface of the target abnormal dimension. By using the present application, the efficiency of detecting abnormal transactions can be improved.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular, to an image processing method, apparatus, computer device, and storage medium. Background Art

[0002] Transactions on the Internet adopt the methods of account verification and certificate verification. As long as the user enters the correct account information, password, and customer certificate, the trading system will recognize the validity of the trading operation. It can be seen that the concealment and particularity of transactions occurring on the network increase the difficulty of detecting suspicious transactions.

[0003] Currently, in order to detect suspicious transactions among a large number of transactions, all transactions are manually verified one by one. During manual verification, based on past experience, suspicious information is discovered from the transactions, and pictures containing the suspicious information are intercepted from the transactions for subsequent review. Due to the huge number of transactions, relying solely on manual discovery of suspicious information in the transactions and manually intercepting evidence pictures will reduce the efficiency of detecting abnormal transactions. Summary of the Invention

[0004] Embodiments of this application provide an image processing method, apparatus, computer device, and storage medium, which can improve the efficiency of detecting abnormal transactions.

[0005] On the one hand, an embodiment of this application provides an image processing method, including:

[0006] Display a transaction monitoring interface, where the transaction monitoring interface includes historical transaction data and an image display control;

[0007] When the image display control is selected, display an abnormal result interface; the abnormal result interface includes one or more abnormal dimensions, and each abnormal dimension includes one or more abnormal transaction data images. The abnormal transaction data images under each abnormal dimension are obtained by detecting the historical transaction data using a detection rule corresponding to each abnormal dimension;

[0008] When there is a preview operation for a target abnormal dimension, display a preview interface for the target abnormal dimension, and display one or more abnormal transaction data images under the target abnormal dimension in the preview interface for the target abnormal dimension. The target abnormal dimension is one of the one or more abnormal dimensions.

[0009] On the one hand, an embodiment of this application provides an image processing apparatus, including:

[0010] A first display module, configured to display a transaction monitoring interface, where the transaction monitoring interface includes historical transaction data and an image display control;

[0011] A second display module, configured to display an abnormal result interface when the image display control is selected; the abnormal result interface includes one or more abnormal dimensions, and each abnormal dimension includes one or more abnormal transaction data images. The abnormal transaction data images under each abnormal dimension are obtained by detecting the historical transaction data using a detection rule corresponding to each abnormal dimension.

[0012] A third display module, configured to display a preview interface of the target abnormal dimension and display one or more abnormal transaction data images under the target abnormal dimension in the preview interface of the target abnormal dimension when there is a preview operation for the target abnormal dimension, where the target abnormal dimension is one of the one or more abnormal dimensions.

[0013] On the one hand, an embodiment of the present application provides a computer device, including a memory and a processor. The memory stores a computer program. When the computer program is executed by the processor, the processor is caused to execute the methods in the above embodiments.

[0014] On the one hand, an embodiment of the present application provides a computer storage medium. The computer storage medium stores a computer program, and the computer program includes program instructions. When the program instructions are executed by a processor, the methods in the above embodiments are executed.

[0015] On the one hand, an embodiment of the present application provides a computer program product or a computer program. The computer program product or the computer program includes computer instructions. The computer instructions are stored in a computer-readable storage medium. When the computer instructions are executed by a processor of a computer device, the methods in the above embodiments are executed.

[0016] By detecting the historical transaction data using detection rules corresponding to multiple abnormal dimensions to obtain abnormal transaction data images under multiple abnormal dimensions, and the user can, through a preview operation on a certain abnormal dimension, display all the abnormal transaction data images under the abnormal dimension on the interface. It can be seen that by automatically identifying suspicious information in the historical transaction data by the terminal device without manual identification, the detection efficiency and detection accuracy of abnormal transaction data can be improved; by detecting the historical transaction data from multiple aspects through multiple abnormal dimensions, the detection accuracy of abnormal transaction data can be further improved; further, through the interface display method, the abnormal transaction data in the historical transaction data under a certain abnormal dimension can be intuitively and clearly displayed, improving the display effect of the abnormal transaction data. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0018] Figure 1 It is a system architecture diagram for image processing provided by an embodiment of the present invention;

[0019] Figures 2a - 2h It is a schematic diagram of an image processing scenario provided by an example of the present application;

[0020] Figure 3 It is a schematic flowchart of an image processing method provided by an embodiment of the present application;

[0021] Figures 4a - 4b It is a schematic interface diagram provided by an embodiment of the present application;

[0022] Figure 5 It is an overall framework diagram for identifying abnormal transaction data provided by an embodiment of the present application;

[0023] Figure 6 It is a schematic flowchart of an image processing method provided by an embodiment of the present application;

[0024] Figure 7 It is a schematic flowchart of a process for determining an abnormal transaction data image provided by an embodiment of the present application;

[0025] Figure 8 It is a schematic flowchart of an image processing method provided by an embodiment of the present application;

[0026] Figure 9 It is a schematic flowchart of a process for determining an abnormal transaction data image provided by an embodiment of the present application;

[0027] Figure 10 It is a schematic structural diagram of an image processing device provided by an embodiment of the present application;

[0028] Figure 11 It is a schematic structural diagram of a computer device provided by an embodiment of the present application. Detailed implementation manners

[0029] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts belong to the scope of protection of the present application.

[0030] Please refer to Figure 1 , which is a system architecture diagram of image processing provided by an embodiment of the present invention. The server 10f is connected to the user terminal cluster through the switch 10e and the communication bus 10d. The user terminal cluster may include: user terminals 10a, 10b,..., user terminal 10c. A large amount of historical transaction data of accounts is stored in the database 10g. For the historical transaction data of an account, the server 10f performs anomaly analysis on the historical transaction data through the detection rules corresponding to multiple anomaly dimensions respectively to discover the abnormal transaction data in the historical transaction data under different anomaly dimensions. Once abnormal transaction data is found, the server generates an image containing the abnormal transaction data. Subsequently, the review personnel can view the images of the abnormal transaction data under each anomaly dimension on the interface in the user terminal.

[0031] It can be seen that by detecting the historical transaction data through multiple anomaly dimensions, the detection accuracy of the historical transaction data can be improved; through the interface display method, the abnormal transaction data in the historical transaction data under a certain anomaly dimension can be intuitively and clearly displayed.

[0032] Figure 1 The terminal devices 10a, 10b, 10c, etc. shown can be intelligent devices with display functions such as mobile phones, tablet computers, laptop computers, palmtop computers, mobile internet devices (MIDs), wearable devices, etc. The terminal device cluster and the server 10f can be directly or indirectly connected through wired or wireless communication methods, and the present application does not make any restrictions here.

[0033] This application can be applied to an abnormal transaction review system (e.g., an anti-money laundering review system). When a reviewer reviews the historical transaction data of a certain account, by clicking on the image display control on the transaction monitoring interface, an abnormal result interface is displayed. This abnormal result interface contains one or more abnormal dimensions, and each abnormal dimension subordinates one or more abnormal transaction data images. Moreover, the abnormal transaction data images under each abnormal dimension are obtained by detecting the historical transaction data using the detection rules corresponding to each abnormal dimension. Generally speaking, the abnormal transaction data images under one abnormal dimension will contain the abnormal transaction data in the historical transaction data under this abnormal dimension. The reviewer can perform a preview operation for the target abnormal dimension to display all the abnormal transaction data images subordinate to the target abnormal dimension.

[0034] The following takes Figures 2a - 2h as an example to specifically illustrate how to generate abnormal transaction data images and how to display them.

[0035] Please refer to Figures 2a - 2h , which is a schematic diagram of an image processing scenario provided by an example of this application. As Figure 2a shown, after the reviewer logs in to the abnormal transaction review system, they enter the main interface 20a of the abnormal transaction review system. The main interface contains cases to be reviewed and reviewed cases. Each case corresponds to the historical transaction data of an account within a certain period (e.g., 3 months). The processing status of the reviewed cases is: processed. Of course, the reviewed cases will also have corresponding review results; on the contrary, the processing status of the cases to be reviewed is: unprocessed, and the review results of the un-reviewed cases are empty.

[0036] The reviewer can select the case for the current review. As Figure 2b described, the terminal display interface 20b will contain the historical transaction data of account "4643" within a period of time, the button "Details", the button "Abnormal", and the button "Normal", where the button "Abnormal" and the button "Normal" are used to submit the review results. Assume that the number of historical transaction data of account "4643" within a period of time is 3, and each historical transaction data contains details such as transaction flow, account nickname, counterparty account nickname, and transaction remarks. Among them, the transaction flow specifically includes transaction time, transaction amount, main transaction account number, and counterparty transaction account number, etc.

[0037] The reviewer clicks the button "Details". As Figure 2cAs shown, the terminal will display interface 20c. It can be seen from interface 20c that there are three types of abnormal dimensions in interface 20c, namely: transaction flow abnormal dimension, nickname abnormal dimension, and note abnormal dimension. And there are multiple images under each abnormal dimension. The images under each abnormal dimension will contain the abnormal transaction data in the historical transaction data under each abnormal dimension. And the abnormal transaction data in the images is obtained by detecting the historical transaction data through the detection rules under each abnormal dimension.

[0038] In other words, the images under the transaction flow abnormal dimension are obtained by detecting the transaction flow in the historical transaction data using the transaction flow abnormal detection rules; the images under the nickname abnormal dimension are obtained by detecting the nicknames in the historical transaction data using the nickname abnormal detection rules; the images under the note abnormal dimension are obtained by detecting the notes in the historical transaction data using the note abnormal detection rules.

[0039] Interface 20c also includes the button "Overall Preview" under each abnormal dimension, and also includes the button "Preview" corresponding to each abnormal transaction data image. As Figure 2c shown, if the reviewer clicks the button "Overall Preview" under the transaction flow abnormal dimension, as Figure 2d shown, the terminal displays interface 20d, and the abnormal transaction data images (which are image 1 and image 2 respectively) under the transaction flow abnormal dimension are displayed in interface 20d. The abnormal transaction data images will contain the abnormal transaction flows identified based on the transaction flow abnormal detection rules.

[0040] The specific process of generating the abnormal transaction data images under the transaction flow abnormal dimension can be:

[0041] The transaction flow abnormal detection rules can be further divided into a transaction amount rule and a transaction time rule. The transaction amount rule can be specifically: if the proportion of the number of transaction pens with a single transaction amount being an integer multiple of ten yuan or one hundred yuan exceeds 30%, then it is determined that these transaction flows are abnormal transaction flows; the transaction time rule can be specifically: if the proportion of the number of transaction pens within a preset time period (the preset time period can be 22:00 - 06:00) exceeds 30%, then it is determined that these transaction flows are abnormal transaction flows.

[0042] Suppose the abnormal transaction flows identified based on the transaction amount rule are: AA and Ee, and a statistical result can also be generated: 2 / 3 = 67%. This statistical result indicates that 2 abnormal transaction flows are identified, and there are a total of 3 abnormal transaction flows, and the proportion of abnormal transaction flows is 67%. Generate image 1 containing the statistical result of the abnormal transaction flows "AA" and "Ee".

[0043] Suppose the abnormal transaction flows identified based on the transaction time rule are also: AA and Ee, and a time distribution diagram can also be generated. Generate image 2 containing the time distribution diagram of the abnormal transaction flows "AA" and "Ee".

[0044] The abnormal transaction data image in interface 20d can intuitively and clearly display risk points, helping the reviewers quickly identify and analyze cases.

[0045] The abnormal transaction data images under the remaining abnormal dimensions can be obtained by detecting the historical transaction data using the corresponding detection rules, which will not be elaborated here.

[0046] As Figure 2e shown, if the user clicks the button "Preview" corresponding to Image 1 under the abnormal dimension of transaction flow in interface 20c, as Figure 2f shown, the terminal displays interface 20e, and Image 1 under the abnormal dimension of transaction flow is shown in interface 20e.

[0047] After the reviewer previews the abnormal transaction data images under all abnormal dimensions, or after viewing the abnormal transaction data images under some abnormal dimensions, as Figure 2g shown, the reviewer can click the button "Abnormal" or the button "Normal" in interface 20b. If the reviewer believes that there are no abnormalities in the historical transaction data of account "4643", or even if there are some abnormalities, they can be ignored, then the reviewer can click the button "Normal". On the contrary, if the reviewer believes that there are abnormalities in the historical transaction data of account "4643" and these abnormalities cannot be ignored, then the reviewer can click the button "Abnormal".

[0048] Suppose the reviewer clicks the button "Abnormal". At this time, the terminal closes all interfaces except the main interface 20a, and in the main interface, the processing status of account "4643" is changed from "unprocessed" to "processed", and the review result of account "4643" is set to: Abnormal.

[0049] Subsequently, the reviewer can select the next case for review from the main interface 20a.

[0050] Among them, for the specific process of displaying the transaction monitoring interface (such as interface 20b in the above embodiment), displaying the abnormal result interface (such as interface 20c in the above embodiment), where the abnormal result interface contains multiple abnormal dimensions (such as the transaction flow abnormal dimension, nickname abnormal dimension, and note abnormal dimension in the above embodiment), and generating the abnormal transaction data images under each abnormal dimension (such as Image 1 and Image 2 under the transaction flow abnormal dimension, Image 1 and Image 2 under the nickname abnormal dimension, and Image 1 and Image 2 under the note abnormal dimension in the above embodiment), reference can be made to the following Figures 3 - 9 corresponding embodiment.

[0051] Please refer to Figure 3 , Figure 3It is a schematic flowchart of an image processing method provided by an embodiment of the present application. This embodiment mainly shows the display method of abnormal transaction data images from a visualization perspective. The image processing method includes the following steps:

[0052] Step S101, display a transaction monitoring interface, where the transaction monitoring interface includes historical transaction data and an image display control.

[0053] Specifically, the terminal displays a case list interface (such as the main interface 20a in the above Figures 2a - 2h corresponding embodiment). The case list interface includes multiple cases and the status of each case. The status of a case can be an unprocessed status or a completed status.

[0054] When there is an audit operation for a target case, the terminal displays a transaction monitoring interface (such as the interface 20b in the above Figures 2a - 2h corresponding embodiment). This transaction monitoring interface contains historical transaction data and an image display control (such as the button "Details" in the interface 20b in the above Figures 2a - 2h corresponding embodiment). The manifestation form of the image display control can be a button, a link, an icon, etc.

[0055] The audit operation can specifically be that the user clicks on the target case. The target case is one of the multiple cases, and there is a corresponding relationship between the target case and the historical transaction data in the transaction monitoring interface.

[0056] The historical transaction data is the transaction data that has occurred for the same account (which can be called the main account) within a period of time. The historical transaction data can include transaction records and transaction texts. The transaction text can be the transaction account name and can also be a transaction note (the transaction note can also be called a transaction postscript, which is used to indicate the purpose of this transaction). The account name can be the user nickname of the user in the application where the transaction is initiated.

[0057] The transaction records can specifically include: transaction time, transaction amount, digital account numbers of both parties to the transaction, etc. The transaction account name can be the account names of both parties to the transaction and can also be the account names of other associated accounts under the main account, etc. The transaction note can be the note text generated by either party when initiating the transaction and can also be the note text generated by the main account or other associated accounts under the main account when the main account conducts transactions with other associated accounts under the main account.

[0058] The number of transaction records can be one or more, the number of transaction account names can be one or more, and the number of transaction notes can be one or more.

[0059] Step S102, when the image display control is selected, display an abnormal result interface; the abnormal result interface includes one or more abnormal dimensions, and each abnormal dimension includes one or more abnormal transaction data images. The abnormal transaction data images under each abnormal dimension are obtained by detecting the historical transaction data using the detection rules corresponding to each abnormal dimension.

[0060] Specifically, when the image display control is selected, display an abnormal result interface (such as interface 20c in the corresponding embodiment above). The abnormal result interface includes one or more abnormal dimensions, and each abnormal dimension includes one or more abnormal transaction data images. The abnormal transaction data images under each abnormal dimension are obtained by detecting the historical transaction data using the detection rules corresponding to each abnormal dimension. Figures 2a - 2h Each abnormal transaction data image contains abnormal transaction data, and the abnormal transaction data in the abnormal transaction data under the abnormal dimension is determined by detecting the historical transaction data using the detection rules corresponding to the abnormal dimension. That is, the abnormal transaction data under the abnormal dimension is the abnormal information in the historical transaction data under the abnormal dimension.

[0061] In addition to including each abnormal dimension, the abnormal result interface also includes a general preview control corresponding to each abnormal dimension (the general preview control is a preview control for multiple abnormal transaction data under one abnormal dimension), and also includes the image name and preview control of each abnormal transaction data image under each abnormal dimension (the preview control of the abnormal transaction data image is a preview control for a single abnormal transaction data image).

[0062] It should be noted that in the abnormal result interface, the image names and preview controls of all abnormal transaction data images under each abnormal dimension are arranged and displayed according to the display priority of the abnormal transaction data images. Of course, the higher the display priority, the more forward the display position of the image name and preview control of the abnormal transaction data image.

[0063] The display priority of the abnormal transaction data images is determined by the risk prediction label of the abnormal transaction data images, and each abnormal transaction data image carries a risk prediction label, which is used to indicate the risk degree of the abnormal transaction data in the abnormal transaction data image.

[0064]

[0065] ​For example, under a certain abnormal dimension, there are abnormal transaction data image 1 and abnormal transaction data image 2. If the risk prediction label of abnormal transaction data image 1 is: high risk, and the risk prediction label of abnormal transaction data image 2 is: low risk, then on the abnormal result interface, the image name and the display position of the preview control of abnormal transaction data image 1 are more prominent than those of abnormal transaction data image 2.

[0066] Step S103, when there is a preview operation for the target abnormal dimension, display the preview interface of the target abnormal dimension, and display one or more abnormal transaction data images under the target abnormal dimension in the preview interface of the target abnormal dimension, where the target abnormal dimension is one of one or more abnormal dimensions.

[0067] Specifically, the target abnormal dimension is one of one or more abnormal dimensions, and the preview operation for the target abnormal dimension may specifically refer to the selection operation of the overall preview control corresponding to the target abnormal dimension.

[0068] Display the preview interface of the target abnormal dimension (such as interface 20d in the above Figures 2a - 2h corresponding embodiment), and display all abnormal transaction data images under the target abnormal dimension in the preview interface of the target abnormal dimension.

[0069] As can be seen from the foregoing, all abnormal transaction data images will carry a risk prediction label, which represents the risk level of the abnormal transaction data in the abnormal transaction data image. Therefore, when displaying the abnormal transaction data images under the target abnormal dimension on the preview interface of the target abnormal dimension, the abnormal transaction data images under the target abnormal dimension can also be arranged and displayed according to the display priority.

[0070] Under the target abnormal dimension, the display priority of each abnormal transaction data image is also determined by the risk prediction label of each abnormal transaction data image.

[0071] Optionally, the transaction monitoring interface further includes an audit completion control (such as the buttons "abnormal" and "normal" in interface 20b in the above Figures 2a - 2h corresponding embodiment). When the user clicks on this audit completion control, it means that the user has completed the audit of the target case, and the transaction monitoring interface, the abnormal result interface, and the preview interface of the target abnormal dimension can be closed. After closing, only the case list interface is displayed on the screen, and the status of the target case is adjusted to the completed status in the case list interface.

[0072] Optionally, as can be seen from the foregoing, the abnormal result interface further includes a preview control for each abnormal transaction data image under each abnormal dimension. Therefore, when the preview control of the target abnormal transaction data image is selected, a preview interface of the target abnormal transaction data image is displayed, and the target abnormal transaction data image is displayed in the preview interface of the target abnormal transaction data image. The target abnormal transaction data image is one of all the abnormal transaction data images under all abnormal dimensions.

[0073] Optionally, the preview interface of the target abnormal transaction data and the preview interface of the target abnormal dimension are different from each other, and both of the above interfaces belong to the floating window interface. Since the floating window interface is a lightweight window, the time consumed for creating and deleting the floating window is short. At the same time, when the floating window interface is displayed, it does not affect the display of the remaining windows. Therefore, the transaction monitoring interface, the preview interface of the target abnormal transaction data, and the preview interface of the target abnormal dimension can be displayed simultaneously, that is, the user can preview the above three windows on the screen at the same time.

[0074] Optionally, the abnormal dimension can be further divided into multiple unit abnormal dimensions. The number of unit abnormal dimensions included in each abnormal dimension can be one or multiple. There is a corresponding detection rule for each unit abnormal dimension. By detecting the historical transaction data based on the detection rule corresponding to the unit abnormal dimension, the abnormal transaction data in the historical transaction data can be obtained. The number of abnormal transaction data images under an abnormal dimension = the sum of the abnormal transaction data images detected based on the detection rules corresponding to all the unit abnormal dimensions under that abnormal dimension.

[0075] Therefore, on the abnormal result interface, in addition to displaying one or more abnormal dimensions, the unit abnormal dimensions included in each abnormal dimension will also be displayed. On the abnormal result interface, the image names and image preview controls of all the abnormal transaction data images under all abnormal dimensions can be displayed together; it is also possible to switch to display the image names and image preview controls under each unit abnormal dimension, that is, when a certain unit abnormal dimension is selected, only the image names and image preview controls of all the abnormal transaction data images under the selected unit abnormal dimension are displayed on the abnormal result interface. When there is a preview operation for the target unit abnormal dimension, a preview interface of the target unit abnormal dimension is displayed, and the abnormal transaction data images under the target unit abnormal dimension are displayed in the preview interface of the target unit abnormal dimension.

[0076] Of course, whether directly displaying the image names and image preview controls of all the abnormal transaction data images under all abnormal dimensions or only displaying the image names and image preview controls of all the abnormal transaction data images under the selected unit abnormal dimension, the image names of the abnormal transaction data images are arranged and displayed according to the display priority of the corresponding abnormal transaction data images.

[0077] Please refer to Figures 4a - 4b , which is a schematic interface diagram provided by an embodiment of the present application. Figure 4a The interface of Figure 4b can correspond to the abnormal result interface in the present application, and the interface of Figure 4a can correspond to the preview interface of the target abnormal dimension in the present application. The transaction flow, nickname, and remarks in Figure 4a can correspond to the abnormal dimensions in the present application. The transaction amount, transaction time, and transaction frequency are three sub-abnormal dimensions under the abnormal dimension of "transaction flow"; the main nickname, associated nickname, and opponent nickname are three sub-abnormal dimensions under the abnormal dimension of "nickname"; the main remarks, associated remarks, and opponent remarks are three sub-abnormal dimensions under the abnormal dimension of "remarks". As can be seen from Figure 4b , the sub-abnormal dimensions under each abnormal dimension can be switched and displayed. The user can preview an abnormal transaction data image, preview an abnormal transaction data image under a certain sub-abnormal dimension, or preview an abnormal transaction data image under a certain abnormal dimension. If the user selects to preview all the abnormal transaction data images under the abnormal dimension of "transaction flow", the interface of

[0078] Please refer to Figure 5 will be displayed, and all the abnormal transaction data images under the abnormal dimension of "transaction flow" will be shown in this interface. Figure 5 is an overall framework diagram for identifying abnormal transaction data provided by an embodiment of the present application. As shown in Figure 5 , the present application involves the initiator and the background. The initiator can be an abnormal transaction audit system for auditing abnormal transaction data, mainly used to display abnormal transaction data for manual auditing; the main function of the background is to identify abnormal transaction data and display it to the initiator (the background can also be embedded in the abnormal transaction audit system). The initiator extracts the un-audited historical transaction data from the transaction pool and generates corresponding cases. When the user selects a case for auditing, the background can extract the historical transaction data corresponding to the case, determine the abnormal transaction data in the historical transaction data under different abnormal dimensions by matching the abnormal indicators of transaction characteristics and the black word library, and generate an abnormal transaction data image containing the abnormal transaction data. The multiple abnormal dimensions can specifically be the transaction flow abnormal dimension, name abnormal dimension, and remarks abnormal dimension. The corresponding abnormal transaction data images are respectively displayed on the interface where the abnormal transaction audit system is located, and when the auditor processes the case, they can directly view the abnormal transaction data images under different abnormal dimensions to improve the audit efficiency.

[0079] This application can actively identify risk points and display abnormal information in the form of images on the front-end page. The online launch of this application saves the operation time of reviewers and greatly improves work efficiency. For example, a case generally requires dozens of suspicious points to support the conclusion. The overall time-consuming for manual review of a case is about 6 minutes, and the time-consuming for taking screenshots of suspicious points is about 2 minutes. The online launch of this application can save 40% of the time. In addition, this application can directly display abnormal information on the task page, providing strong evidence for reviewers' case analysis, making the output of suspicious transaction messages more substantial, and thus improving the review quality and accuracy. In summary, while improving the work efficiency of reviewers, it can ensure the accuracy of case determination and also achieve the purpose of saving manpower.

[0080] Please refer to Figure 6 , Figure 6 which is a schematic flowchart of an image processing method provided by an embodiment of this application. This embodiment mainly describes how to determine abnormal transaction data images from historical transaction data. The image processing method includes the following steps:

[0081] Step S201: Extract the unit historical transaction data corresponding to each abnormal dimension from the historical transaction data.

[0082] Specifically, the terminal obtains the historical transaction data to be detected. Among them, the historical transaction data is the transaction data that has occurred for the same account (which can be called the main account) within a period of time. The historical transaction data can include transaction records and transaction texts. The transaction text can be the transaction account name and can also be the transaction note (the transaction note can also be called the transaction postscript, which is used to indicate the purpose of this transaction). The account name can be the user nickname of the user in the application where the transaction is initiated.

[0083] The transaction records can specifically include: transaction time, transaction amount, digital account numbers of both parties to the transaction, etc. The transaction account name can be the account names of both parties to the transaction and can also be the account names of other associated accounts under the main account, etc. The transaction note can be the note text generated when any party to the transaction initiates the transaction, and can also be the note text generated by the main account or other associated accounts under the main account when the main account conducts transactions with other associated accounts under the main account.

[0084] The number of transaction records can be one or more, the number of transaction account names can be one or more, and the number of transaction notes can be one or more.

[0085] The terminal obtains multiple anomaly dimensions, which can be divided according to the processing object. The multiple anomaly dimensions can be classified into transaction flow anomaly dimensions and text anomaly dimensions. The terminal extracts the unit historical transaction data corresponding to the transaction flow anomaly dimension from the historical transaction data. Here, the unit historical transaction data corresponding to the transaction flow anomaly dimension is the transaction flow in the historical transaction data. The terminal extracts the unit historical transaction data corresponding to the text anomaly dimension from the historical transaction data. Here, the unit historical transaction data corresponding to the text anomaly dimension is the transaction text in the historical transaction data.

[0086] Step S202, determine the detection rules corresponding to each anomaly dimension.

[0087] Specifically, the terminal obtains the detection rules corresponding to each anomaly dimension. The detection rules are set manually based on experience, and the detection rules will define the detection object and the detection means.

[0088] Step S203, detect the unit historical transaction data corresponding to each anomaly dimension according to the detection rules corresponding to each anomaly dimension, and obtain the anomaly transaction data image under each anomaly dimension.

[0089] Specifically, as can be seen from the above, the anomaly dimensions are divided into transaction flow anomaly dimensions and text anomaly dimensions. Below, the unit historical transaction data under these two anomaly dimensions are detected respectively to obtain the anomaly transaction data image.

[0090] First, explain how to determine the anomaly transaction data image under the transaction flow anomaly dimension:

[0091] The target anomaly dimension is the transaction flow anomaly dimension. The unit historical transaction data corresponding to the target anomaly dimension is the transaction flow, and the number of transaction flows is multiple. Detect the multiple transaction flows according to the detection rules corresponding to the target anomaly dimension to obtain the abnormal transaction flows in the multiple transaction flows. Generate a statistical analysis result based on the abnormal transaction flows, the multiple transaction flows, and the detection rules corresponding to the target anomaly dimension. The terminal generates an image containing the above abnormal transaction flows and the statistical analysis result. This image is the anomaly transaction data image under the target anomaly dimension, and the abnormal transaction data in this anomaly transaction data image is the abnormal transaction flow.

[0092] Furthermore, the transaction flow anomaly dimension can be subdivided into multiple unit anomaly dimensions, and each unit anomaly dimension corresponds to a detection rule. The terminal can determine the abnormal transaction flows under different unit anomaly dimensions according to the detection rules corresponding to each unit anomaly dimension respectively. Similarly, generate the statistical analysis result under each unit anomaly dimension and generate the anomaly transaction data image under each unit anomaly dimension respectively.

[0093] For example, the abnormal transaction volume dimension includes 3 unit abnormal dimensions, namely the amount abnormal dimension, the time abnormal dimension, and the frequency abnormal dimension. The detection rule for the amount abnormal dimension can be specifically: if the proportion of the number of transaction records with the amount being an integer multiple of ten yuan or one hundred yuan in a single transaction record exceeds a threshold (e.g., 30%), then the transaction records with the amount being an integer multiple of ten yuan or one hundred yuan are determined as abnormal transaction records. The statistical analysis result for the amount abnormal dimension can be the statistical proportion of abnormal transaction records to all transaction records.

[0094] The detection rule for the time abnormal dimension can be specifically: if the proportion of the number of transaction records with the transaction time within a preset time period exceeds a threshold (e.g., 30%) (the value of the preset time period can be from 22:00 to 06:00), then the transaction records with the transaction time within the preset time period are determined as abnormal transaction records. The statistical analysis result for the time abnormal dimension can be a time distribution graph of abnormal transaction records and the remaining transaction records in the transaction records except for the abnormal transaction records.

[0095] The detection process for the frequency abnormal dimension can be specifically: if the proportion of the cases where the transaction frequency reaches a frequency threshold (e.g., 60 times) within a time period (e.g., within one hour) exceeds a percentage threshold (e.g., 30%), then the transaction records with the transaction frequency reaching the frequency threshold within the time period are regarded as abnormal transaction records. The statistical analysis result for the frequency abnormal dimension can be a trend graph of the transaction frequency in each time period, and the time period where the abnormal transaction records are located is marked in the trend graph.

[0096] As can be seen from the above, the detection objects of the 3 detection rules corresponding to the 3 unit abnormal dimensions under the abnormal transaction volume dimension are all multiple transaction records.

[0097] Optionally, when determining the abnormal transaction data image under each unit abnormal dimension, the statistical analysis result is included. According to the statistical analysis result, a risk prediction label for the abnormal transaction record is determined, and an association relationship is set between the risk prediction label and the abnormal transaction data image containing the abnormal transaction record. The risk prediction label can be used for the display priority of the abnormal transaction data image.

[0098] For example, when the abnormal transaction data image is determined by the detection rule corresponding to the amount abnormal dimension, the higher the proportion of the number of transaction records with the amount being an integer multiple of ten yuan or one hundred yuan in a single transaction record, the higher the risk level corresponding to the risk prediction label of the abnormal transaction data image;

[0099] when the abnormal transaction data image is determined by the detection rule corresponding to the time abnormal dimension, the higher the proportion of the number of transaction records with the transaction time within the preset time period, the higher the risk level corresponding to the risk prediction label of the abnormal transaction data image;

[0100] When the abnormal transaction data image is determined by the detection rule corresponding to the frequency anomaly dimension, the higher the proportion of transactions reaching the frequency threshold within the time period, the higher the risk level corresponding to the risk prediction label of the abnormal transaction data image.

[0101] Next, it will be described how to determine the abnormal transaction data image under the text anomaly dimension:

[0102] The target anomaly dimension is the text anomaly dimension. The unit historical transaction data corresponding to the text anomaly dimension includes transaction texts, and the number of transaction texts is multiple. The transaction texts can be transaction account names or transaction remarks.

[0103] According to the detection rule corresponding to the target anomaly dimension, the abnormal keyword library is respectively matched with each transaction text to obtain abnormal keywords; the transaction text where the abnormal keyword is located is used as the abnormal transaction text, and the risk prediction label corresponding to the abnormal keyword is obtained, and an image containing the abnormal transaction text and the risk prediction label is generated, and this image is the abnormal transaction data image under the text anomaly dimension, and the abnormal transaction data in this abnormal transaction data image is the abnormal transaction text.

[0104] The abnormal keywords are marked in the abnormal transaction data image according to the preset marking method. For example, the preset marking method is to bold, or add a rectangular box, or use different colors, etc.

[0105] Please refer to Figure 7 , Figure 7 FIG. is a schematic flowchart of a process for determining an abnormal transaction data image provided by an embodiment of the present application. The abnormal transaction review system extracts transaction flows from historical transaction data, and detects the transaction flows through the detection rules corresponding to the flow anomaly dimension to extract abnormal transaction flows and statistical analysis results. Among them, the detection rules corresponding to the flow anomaly dimension may include detection rules related to transaction amounts, detection rules related to transaction times, and detection rules related to transaction frequencies. These 3 detection rules here can correspond to the 3 detection rules respectively corresponding to the 3 unit anomaly dimensions under the flow anomaly dimension in the present application. An abnormal transaction data image containing the abnormal transaction flow and the statistical analysis result is generated, and the above abnormal transaction data image is displayed under the flow anomaly dimension on the front-end page.

[0106] Please refer to Figure 8 , Figure 8 FIG. is a schematic flowchart of an image processing method provided by an embodiment of the present application. This embodiment mainly describes how to determine the abnormal transaction data image under the text anomaly dimension when the target anomaly dimension is the text anomaly dimension. The image processing method includes the following steps:

[0107] Step S301: According to the detection rules corresponding to the target abnormal dimension, match the abnormal keyword library with each transaction text to obtain abnormal keywords.

[0108] Step S302: Take the transaction text where the abnormal keyword is located as the abnormal transaction text.

[0109] The text abnormal dimension can be the name abnormal dimension or the note abnormal dimension, and the above two abnormal dimensions both belong to the abnormal dimensions in one or more abnormal dimensions. The processing object of the name abnormal dimension is the transaction account name, and the processing object of the note abnormal dimension is the transaction note.

[0110] When the text abnormal dimension is the name abnormal dimension, the transaction text is the transaction account name, the abnormal transaction text is the abnormal account name, and the number of transaction account names is multiple. Further, the name abnormal dimension can be divided into multiple unit abnormal dimensions, and each unit abnormal dimension corresponds to a detection rule. The terminal can determine the abnormal account names under different unit abnormal dimensions according to the detection rules corresponding to each unit abnormal dimension respectively.

[0111] As can be seen from the above, the account names of both parties to the transaction (the account names of both parties to the transaction include the main account name and the counterparty account name), as well as the account names of other associated accounts under the main account are all called transaction account names.

[0112] Divide multiple transaction account names into sets of transaction account names belonging to different unit abnormal dimensions, and use the detection rules corresponding to different unit abnormal dimensions to determine abnormal keywords and the abnormal transaction account names where the abnormal keywords are located from the corresponding sets of transaction account names.

[0113] For example, if the name abnormal dimension includes 3 unit abnormal dimensions, namely the main account name abnormal dimension, the counterparty account name abnormal dimension, and the associated account name abnormal dimension, then correspondingly, the main account names can be combined into a set corresponding to the main account name abnormal dimension, all the counterparty account names can be combined into a set corresponding to the counterparty account name abnormal dimension, and all the associated account names can be combined into a set corresponding to the associated account name abnormal dimension.

[0114] The detection rule corresponding to the main account name abnormal dimension can be specifically: if the main account name matches successfully with the abnormal keyword library, then take the main account name as the abnormal account name; successful matching means that there is at least one phrase that exists both in the main account name and in the abnormal keyword library, and take this phrase as the abnormal keyword of the abnormal account name.

[0115] The detection rule corresponding to the abnormal dimension of the opponent's account name can be specifically: if the opponent's account name matches the abnormal keyword library, then the opponent's account name is regarded as an abnormal account name; a successful match means that there is at least one phrase that exists in both the opponent's account name and the abnormal keyword library, and this phrase is used as the abnormal keyword of the abnormal account name.

[0116] The detection rule corresponding to the abnormal dimension of the associated account name can be specifically: if the associated account name matches the abnormal keyword library, then the associated account name is regarded as an abnormal account name; a successful match means that there is at least one phrase that exists in both the associated account name and the abnormal keyword library, and this phrase is used as the abnormal keyword of the abnormal account name.

[0117] It can be known that the detection object of the detection rule corresponding to the abnormal dimension of the main account name is the main account name, the detection object of the detection rule corresponding to the abnormal dimension of the opponent's account name is the opponent's account name, and the detection object of the detection rule corresponding to the abnormal dimension of the associated account name is the associated account name.

[0118] Step S303, generate the abnormal transaction data image under the target abnormal dimension, and the abnormal transaction data image under the target abnormal dimension includes the abnormal transaction text.

[0119] Similarly, determine the risk prediction label of each abnormal keyword, and generate the abnormal transaction data image including the abnormal account name and the risk prediction label under each unit abnormal dimension. The generated abnormal transaction data image is the abnormal transaction data image under the name abnormal dimension.

[0120] Optionally, when the text abnormal dimension is the note abnormal dimension, the transaction text is the transaction note, the abnormal transaction text is the abnormal transaction note, and the number of transaction notes is multiple, furthermore, the note abnormal dimension can be subdivided into multiple unit abnormal dimensions, and each unit abnormal dimension corresponds to a detection rule. The terminal can respectively determine the abnormal transaction notes under different unit abnormal dimensions according to the detection rules corresponding to each unit abnormal dimension. As can be seen from the foregoing, the note text generated when any one of the trading parties initiates a transaction and the note text generated by the main account or other associated accounts under the main account when the main account conducts a transaction with other associated accounts under the main account are all called transaction notes.

[0121] Divide multiple transaction notes into transaction note sets belonging to different unit abnormal dimensions, and use the detection rules corresponding to different unit abnormal dimensions to determine the abnormal keywords and the abnormal transaction notes where the abnormal keywords are located from the corresponding transaction note sets.

[0122] For example, the note exception dimension includes 3 unit exception dimensions, namely the main account note exception dimension, the counterparty account note exception dimension, and the associated account note exception dimension. Correspondingly, the transaction notes initiated by the main account can be divided into a set corresponding to the main account note exception dimension, the transaction notes initiated by the counterparty account can be divided into a set corresponding to the counterparty account note exception dimension, and the transaction notes initiated by the associated account can be divided into a set corresponding to the associated account note exception dimension.

[0123] The detection rule corresponding to the main account note exception dimension can be specifically: if the note text initiated by the main account matches the exception keyword library successfully, then the note text is regarded as an abnormal transaction note; successful matching means that there is at least one phrase that exists both in the note text initiated by the main account and in the exception keyword library, and this phrase is used as the exception keyword of this abnormal transaction note.

[0124] The detection rule corresponding to the counterparty account note exception dimension can be specifically: if the note text initiated by the counterparty account matches the exception keyword library successfully, then the note text is regarded as an abnormal transaction note; successful matching means that there is at least one phrase that exists both in the note text initiated by the counterparty account and in the exception keyword library, and this phrase is used as the exception keyword of this abnormal transaction note.

[0125] The detection rule corresponding to the associated account note exception dimension can be specifically: if the note text initiated by the associated account matches the exception keyword library successfully, then the note text is regarded as an abnormal transaction note; successful matching means that there is at least one phrase that exists both in the note text initiated by the associated account and in the exception keyword library, and this phrase is used as the exception keyword of this abnormal transaction note.

[0126] It can be known that the detection object of the detection rule corresponding to the main account note exception dimension is the note text initiated by the main account, the detection object of the detection rule corresponding to the counterparty account note exception dimension is the note text initiated by the counterparty account, and the detection object of the detection rule corresponding to the associated account note exception dimension is the note text initiated by the associated account.

[0127] Similarly, determine the risk prediction label of each exception keyword, and generate an abnormal transaction data image including abnormal transaction notes and risk prediction labels under each unit exception dimension.

[0128] Generally speaking, this application proposes 3 types of exception dimensions, namely the transaction flow exception dimension, the name exception dimension, and the note exception dimension. The name exception dimension and the note exception dimension can also be called text exception dimensions. Determine the abnormal transaction data and abnormal transaction data images under these 3 types of exception dimensions respectively.

[0129] Please refer toFigure 9 , Figure 9 It is a flow chart of determining an abnormal transaction data image provided by an embodiment of the present application. The abnormal transaction review system extracts account nicknames and transaction notes from historical transaction data, and matches the extracted account nicknames and transaction notes respectively through a black word library to find abnormal keywords in the account nicknames and abnormal keywords in the transaction notes. Account nicknames with abnormal keywords are extracted, and an image of the account nickname containing the abnormal keywords is generated. The image is the abnormal transaction data image, which is displayed under the name abnormal dimension of the front-end page. Transaction notes with abnormal keywords are extracted, and an image of the transaction notes containing abnormal keywords is generated. The image is the abnormal transaction data image, which is displayed under the note abnormal dimension of the front-end page.

[0130] From the above, it can be seen that the present application analyzes historical transaction data from multiple dimensions such as transaction flow and transaction text to discover abnormal information in the historical transaction data. The detection scope is comprehensive and the accuracy of detecting abnormal information can be improved.

[0131] For further information, see Figure 10 , which is a schematic diagram of the structure of an image processing device provided in an embodiment of the present application. Figure 10 As shown, the image processing device 1 can be applied to the above Figures 3 - 9 The terminal in the corresponding embodiment. Specifically, the image processing device 1 can be a computer program (including program code) running in a computer device, for example, the image processing device 1 is an application software; the image processing device 1 can be used to execute the corresponding steps in the method provided in the embodiment of the present application.

[0132] The image processing device 1 may include: a first display module 11 , a second display module 12 and a third display module 13 .

[0133] A first display module 11, used to display a transaction monitoring interface, wherein the transaction monitoring interface includes historical transaction data and image display controls;

[0134] The second display module 12 is used to display an abnormal result interface when the image display control is selected; the abnormal result interface includes one or more abnormal dimensions, each abnormal dimension includes one or more abnormal transaction data images, and the abnormal transaction data image under each abnormal dimension is obtained by detecting the historical transaction data using the detection rules corresponding to each abnormal dimension;

[0135] The third display module 13 is configured to display a preview interface of the target exception dimension and display one or more abnormal transaction data images under the target exception dimension in the preview interface of the target exception dimension when there is a preview operation for the target exception dimension, where the target exception dimension is one of one or more exception dimensions.

[0136] In a possible implementation, the exception result interface further includes a preview control for each abnormal transaction data image under each exception dimension;

[0137] The image processing device 1 may further include: a fourth display module 14.

[0138] The fourth display module 14 is configured to display a preview interface of the target abnormal transaction data image and display the target abnormal transaction data image in the preview interface of the target abnormal transaction data image when the preview control of the target abnormal transaction data image is selected, where the target abnormal transaction data image is one of all abnormal transaction data images under all exception dimensions.

[0139] In a possible implementation, the preview interface of the target abnormal transaction data image and the preview interface of the target exception dimension are different from each other, and both the preview interface of the target abnormal transaction data image and the preview interface of the target exception dimension belong to floating window interfaces.

[0140] In a possible implementation, each abnormal transaction data image under the target exception dimension carries a risk prediction label, and the risk prediction label represents the risk degree of the abnormal transaction data in the abnormal transaction data image;

[0141] When the third display module 13 is configured to display one or more abnormal transaction data images under the target exception dimension in the preview interface of the target exception dimension, it is specifically configured to:

[0142] Arrange and display one or more abnormal transaction data images under the target exception dimension in the preview interface of the target exception dimension according to the display priority, where the display priority of one or more abnormal transaction data images under the target exception dimension is determined by the risk prediction labels of one or more abnormal transaction data images under the target exception dimension.

[0143] In a possible implementation, when the first display module 11 is configured to display the transaction monitoring interface, it is specifically configured to:

[0144] Display a case list interface, where the case list interface includes multiple cases and the status of each case;

[0145] When there is an audit operation for the target case, display the transaction monitoring interface;

[0146] The transaction monitoring interface further includes an audit completion control, and the image processing device 1 may further include: a closing module 15.

[0147] The closing module 15 is configured to close the transaction monitoring interface, the abnormal result interface, and the preview interface of the target abnormal dimension when the audit completion control is selected, and adjust the status of the target case to the completed status in the case list interface.

[0148] In a possible implementation, the image processing device 1 may further include: an extraction module 16 and a detection module 17.

[0149] The extraction module 16 is configured to extract the unit historical transaction data corresponding to each abnormal dimension from the historical transaction data, and determine the detection rules respectively corresponding to each abnormal dimension;

[0150] The detection module 17 is configured to detect the unit historical transaction data corresponding to each abnormal dimension according to the detection rules corresponding to each abnormal dimension, and obtain the abnormal transaction data image under each abnormal dimension.

[0151] In a possible implementation, the unit historical transaction data corresponding to the target abnormal dimension includes transaction flows, and the number of transaction flows is multiple;

[0152] When the detection module 17 is configured to detect the unit historical transaction data corresponding to the target abnormal dimension according to the detection rules corresponding to the target abnormal dimension and obtain the abnormal transaction data image under the target abnormal dimension, it is specifically configured to:

[0153] Detect the multiple transaction flows according to the detection rules corresponding to the target abnormal dimension, and obtain the abnormal transaction flows in the multiple transaction flows;

[0154] Generate the abnormal transaction data image under the target abnormal dimension, and the abnormal transaction data image under the target abnormal dimension includes the abnormal transaction flows.

[0155] In a possible implementation, when the detection module 17 is configured to generate the abnormal transaction data image under the target abnormal dimension, it is specifically configured to:

[0156] Generate a statistical analysis result according to the abnormal transaction flows, the multiple transaction flows, and the detection rules corresponding to the target abnormal dimension;

[0157] Generate the abnormal transaction data image under the target abnormal dimension, and the abnormal transaction data image under the target abnormal dimension includes the abnormal transaction flows and the statistical analysis result.

[0158] In a possible implementation, the image processing device 1 may further include: a statistics module 18.

[0159] The statistics module 18 is configured to determine a risk prediction label for the abnormal transaction flow according to the statistical analysis result, and set an association relationship between the risk prediction label of the abnormal transaction flow and the abnormal transaction data image under the target abnormal dimension.

[0160] In a possible implementation, the unit historical transaction data corresponding to the target abnormal dimension includes transaction texts, and the number of the transaction texts is multiple;

[0161] When the detection module 17 is used to detect the unit historical transaction data corresponding to the target abnormal dimension according to the detection rule corresponding to the target abnormal dimension to obtain the abnormal transaction data image under the target abnormal dimension, it is specifically configured to:

[0162] Detect the multiple transaction texts according to the detection rule corresponding to the target abnormal dimension to obtain the abnormal transaction texts in the multiple transaction texts;

[0163] Generate the abnormal transaction data image under the target abnormal dimension, where the abnormal transaction data image under the target abnormal dimension includes the abnormal transaction texts.

[0164] In a possible implementation, when the detection module 17 is used to detect the multiple transaction texts according to the detection rule corresponding to the target abnormal dimension to obtain the abnormal transaction texts in the multiple transaction texts, it is specifically configured to:

[0165] Match the abnormal keyword library with each transaction text according to the detection rule corresponding to the target abnormal dimension to obtain abnormal keywords;

[0166] Use the transaction text where the abnormal keyword is located as the abnormal transaction text.

[0167] In a possible implementation, when the detection module 17 is used to generate the abnormal transaction data image under the target abnormal dimension, it is specifically configured to:

[0168] Obtain the risk prediction label corresponding to the abnormal keyword;

[0169] Generate the abnormal transaction data image under the target abnormal dimension, where the abnormal transaction data image under the target abnormal dimension includes the abnormal transaction texts and the risk prediction label corresponding to the abnormal keyword, and the abnormal keyword is marked in the abnormal transaction data image under the target abnormal dimension according to a preset marking method.

[0170] According to an embodiment of the present invention, Figures 3 - 9Each step involved in the method shown can be performed by Figure 10 each module in the image processing device shown. For example, Figure 3 the steps S101 - S103 shown in Figure 10 can be performed by the first display module 11, the second display module 12, the third display module 13, the fourth display module 14, and the shutdown module 15 shown in Figure 6 respectively; for another example, Figure 10 the steps S201 - S204 shown in Figure 8 can be performed by the extraction module 16, the detection module 17, and the statistics module 18 shown in Figure 10 ; for another example,

[0171] the steps S301 - S303 shown in Figure 11 can be performed by the detection module 17 shown in Figures 3 - 9 . Figure 11 As shown, the computer device 1000 may include: a user interface 1002, a processor 1004, an encoder 1006, and a memory 1008. The signal receiver 1016 is used to receive or send data via a cellular interface 1010, a WIFI interface 1012,..., or an NFC interface 1014. The encoder 1006 encodes the received data into a data format that can be processed by a computer. A computer program is stored in the memory 1008, and the processor 1004 is configured to execute the steps in any of the above method embodiments through the computer program. The memory 1008 may include a volatile memory (e.g., dynamic random access memory DRAM), and may also include a non-volatile memory (e.g., one-time programmable read-only memory OTPROM). In some instances, the memory 1008 may further include a memory remotely set relative to the processor 1004, and these remote memories can be connected to the computer device 1000 through a network. The user interface 1002 may include: a keyboard 1018 and a display 1020.

[0172] In Figure 11 the computer device 1000 shown, the processor 1004 may be used to call the computer program stored in the memory 1008 to implement:

[0173] display a transaction monitoring interface, the transaction monitoring interface including historical transaction data and an image display control;

[0174] When the image display control is selected, an abnormal result interface is displayed; the abnormal result interface includes one or more abnormal dimensions, and each abnormal dimension includes one or more abnormal transaction data images. The abnormal transaction data images under each abnormal dimension are obtained by detecting the historical transaction data using detection rules corresponding to each abnormal dimension.

[0175] When there is a preview operation for a target abnormal dimension, a preview interface of the target abnormal dimension is displayed, and one or more abnormal transaction data images under the target abnormal dimension are displayed in the preview interface of the target abnormal dimension. The target abnormal dimension is one of the one or more abnormal dimensions.

[0176] In one embodiment, the abnormal result interface further includes preview controls for each abnormal transaction data image under each abnormal dimension.

[0177] The processor 1004 also performs the following steps:

[0178] When the preview control of the target abnormal transaction data image is selected, a preview interface of the target abnormal transaction data image is displayed, and the target abnormal transaction data image is displayed in the preview interface of the target abnormal transaction data image. The target abnormal transaction data image is one of all the abnormal transaction data images under all abnormal dimensions.

[0179] In one embodiment, the preview interface of the target abnormal transaction data image and the preview interface of the target abnormal dimension are different from each other, and both the preview interface of the target abnormal transaction data image and the preview interface of the target abnormal dimension belong to floating window interfaces.

[0180] In one embodiment, each abnormal transaction data image under the target abnormal dimension carries a risk prediction label, and the risk prediction label indicates the risk level of the abnormal transaction data in the abnormal transaction data image.

[0181] When the processor 1004 executes to display one or more abnormal transaction data images under the target abnormal dimension in the preview interface of the target abnormal dimension, it specifically performs the following steps:

[0182] One or more abnormal transaction data images under the target abnormal dimension are arranged and displayed in the preview interface of the target abnormal dimension according to the display priority. The display priority of one or more abnormal transaction data images under the target abnormal dimension is determined by the risk prediction labels of one or more abnormal transaction data images under the target abnormal dimension.

[0183] In one embodiment, when the processor 1004 executes to display the transaction monitoring interface, it specifically performs the following steps:

[0184] Display the case list interface, where the case list interface includes multiple cases and the status of each case;

[0185] When there is an audit operation for the target case, display the transaction monitoring interface;

[0186] The transaction monitoring interface further includes an audit completion control, and the processor 1004 further performs the following steps:

[0187] When the audit completion control is selected, close the transaction monitoring interface, the abnormal result interface, and the preview interface of the target abnormal dimension, and adjust the status of the target case to the completed status in the case list interface.

[0188] In one embodiment, the processor 1004 further performs the following steps:

[0189] Extract the unit historical transaction data corresponding to each abnormal dimension from the historical transaction data;

[0190] Determine the detection rules corresponding to each abnormal dimension;

[0191] Detect the unit historical transaction data corresponding to each abnormal dimension according to the detection rules corresponding to each abnormal dimension, and obtain the abnormal transaction data image under each abnormal dimension.

[0192] In one embodiment, the unit historical transaction data corresponding to the target abnormal dimension includes transaction flows, and the number of transaction flows is multiple;

[0193] When the processor 1004 performs detecting the unit historical transaction data corresponding to the target abnormal dimension according to the detection rule corresponding to the target abnormal dimension to obtain the abnormal transaction data image under the target abnormal dimension, it specifically performs the following steps:

[0194] Detect the multiple transaction flows according to the detection rule corresponding to the target abnormal dimension to obtain the abnormal transaction flows in the multiple transaction flows;

[0195] Generate the abnormal transaction data image under the target abnormal dimension, where the abnormal transaction data image under the target abnormal dimension includes the abnormal transaction flows.

[0196] In one embodiment, when the processor 1004 performs generating the abnormal transaction data image under the target abnormal dimension, it specifically performs the following steps:

[0197] Generate a statistical analysis result according to the abnormal transaction flows, the multiple transaction flows, and the detection rule corresponding to the target abnormal dimension;

[0198] Generate an abnormal transaction data image under the target abnormal dimension, where the abnormal transaction data image under the target abnormal dimension includes the abnormal transaction flow and the statistical analysis result.

[0199] In one embodiment, the processor 1004 further performs the following steps:

[0200] Determine a risk prediction label for the abnormal transaction flow according to the statistical analysis result;

[0201] Set an association relationship between the risk prediction label of the abnormal transaction flow and the abnormal transaction data image under the target abnormal dimension.

[0202] In one embodiment, the unit historical transaction data corresponding to the target abnormal dimension includes transaction texts, and the number of the transaction texts is multiple;

[0203] When the processor 1004 performs detection on the unit historical transaction data corresponding to the target abnormal dimension according to the detection rule corresponding to the target abnormal dimension to obtain an abnormal transaction data image under the target abnormal dimension, the following steps are specifically performed:

[0204] Detect the multiple transaction texts according to the detection rule corresponding to the target abnormal dimension to obtain abnormal transaction texts in the multiple transaction texts;

[0205] Generate an abnormal transaction data image under the target abnormal dimension, where the abnormal transaction data image under the target abnormal dimension includes the abnormal transaction texts.

[0206] In one embodiment, when the processor 1004 performs detection on the multiple transaction texts according to the detection rule corresponding to the target abnormal dimension to obtain abnormal transaction texts in the multiple transaction texts, the following steps are specifically performed:

[0207] Match the abnormal keyword library with each transaction text according to the detection rule corresponding to the target abnormal dimension to obtain abnormal keywords;

[0208] Use the transaction text where the abnormal keyword is located as the abnormal transaction text.

[0209] In one embodiment, when the processor 1004 performs generation of the abnormal transaction data image under the target abnormal dimension, the following steps are specifically performed:

[0210] Obtain the risk prediction label corresponding to the abnormal keyword;

[0211] Generate an abnormal transaction data image under the target abnormal dimension, where the abnormal transaction data image under the target abnormal dimension includes the abnormal transaction text and the risk prediction label corresponding to the abnormal keyword, and the abnormal keyword is marked in the abnormal transaction data image under the target abnormal dimension according to a preset marking method.

[0212] It should be understood that the computer device 1000 described in the embodiments of the present application can execute the description of the image processing method in the corresponding embodiments mentioned above, and can also execute the description of the image processing device 1 in the corresponding embodiments mentioned above, which will not be elaborated here. In addition, the description of the beneficial effects of using the same method will not be elaborated either. Figures 3 - 9 The description of the image processing method in the corresponding embodiments mentioned above, and can also execute the description of the image processing device 1 in the corresponding embodiments mentioned above, which will not be elaborated here. In addition, the description of the beneficial effects of using the same method will not be elaborated either. Figure 10 For the beneficial effects of using the same method, no further description will be given.

[0213] In addition, it should be noted here that: the embodiments of the present application also provide a computer storage medium, and the computer storage medium stores the computer program executed by the image processing device 1 mentioned above, and the computer program includes program instructions. When the processor executes the program instructions, it can execute the description of the image processing method in the corresponding embodiments mentioned above. Therefore, no further description will be given here. In addition, the description of the beneficial effects of using the same method will not be elaborated either. For the technical details not disclosed in the embodiments of the computer storage medium involved in the present application, please refer to the description of the method embodiments of the present application. As an example, the program instructions can be deployed to be executed on one computer device, or on multiple computer devices located at one place, or distributed on multiple computer devices located at multiple places and interconnected through a communication network. The multiple computer devices distributed at multiple places and interconnected through a communication network can be combined into a blockchain network. Figures 3 - 9 For the beneficial effects of using the same method, no further description will be given. For the technical details not disclosed in the embodiments of the computer storage medium involved in the present application, please refer to the description of the method embodiments of the present application. As an example, the program instructions can be deployed to be executed on one computer device, or on multiple computer devices located at one place, or distributed on multiple computer devices located at multiple places and interconnected through a communication network. The multiple computer devices distributed at multiple places and interconnected through a communication network can be combined into a blockchain network.

[0214] According to one aspect of the present application, there is provided a computer program product or a computer program, the computer program product or the computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device can execute the method in the corresponding embodiments mentioned above. Therefore, no further description will be given here. Figures 3 to 9 For the beneficial effects of using the same method, no further description will be given.

[0215] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The above program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above various methods. Among them, the above storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM), etc.

[0216] The above-disclosed are only the preferred embodiments of the present application. Of course, the scope of rights of the present application cannot be limited thereby. Therefore, equivalent changes made according to the claims of the present application still fall within the scope covered by the present application.

Claims

1. An image processing method, characterized in that, Including: Display a transaction monitoring interface, where the transaction monitoring interface includes historical transaction data and an image display control; When the image display control is selected, display an abnormal result interface; the abnormal result interface contains one or more abnormal dimensions, and each abnormal dimension contains one or more abnormal transaction data images. The abnormal transaction data images under each abnormal dimension are obtained by detecting the historical transaction data using a detection rule corresponding to each abnormal dimension; When there is a preview operation for a target abnormal dimension, display a preview interface of the target abnormal dimension, and display one or more abnormal transaction data images under the target abnormal dimension in the preview interface of the target abnormal dimension. The target abnormal dimension is one of the one or more abnormal dimensions.

2. The method according to claim 1, wherein The abnormal result interface further includes a preview control for each abnormal transaction data image under each abnormal dimension; The method further includes; When the preview control of the target abnormal transaction data image is selected, display a preview interface of the target abnormal transaction data image, and display the target abnormal transaction data image in the preview interface of the target abnormal transaction data image. The target abnormal transaction data image is one of the abnormal transaction data images of all abnormal dimensions under all abnormal dimensions.

3. The method according to claim 2, wherein The preview interface of the target abnormal transaction data image and the preview interface of the target abnormal dimension are different from each other, and both the preview interface of the target abnormal transaction data image and the preview interface of the target abnormal dimension belong to floating window interfaces.

4. The method according to claim 1, wherein Each abnormal transaction data image under the target abnormal dimension carries a risk prediction label, and the risk prediction label represents the risk degree of the abnormal transaction data in the abnormal transaction data image; The displaying one or more abnormal transaction data images under the target abnormal dimension in the preview interface of the target abnormal dimension includes: Arranging and displaying one or more abnormal transaction data images under the target abnormal dimension in the preview interface of the target abnormal dimension according to the display priority. The display priority of one or more abnormal transaction data images under the target abnormal dimension is determined by the risk prediction label of one or more abnormal transaction data images under the target abnormal dimension.

5. The method according to any one of claims 1-4, characterized in that, The displaying the transaction monitoring interface includes: Display a case list interface, where the case list interface includes multiple cases and the status of each case; When there is an audit operation for a target case, display the transaction monitoring interface; The transaction monitoring interface further includes an audit completion control, and the method further includes: When the audit completion control is selected, close the transaction monitoring interface, the abnormal result interface, and the preview interface of the target abnormal dimension, and adjust the status of the target case to the completed status in the case list interface.

6. The method according to claim 1, characterized in that, Also including: Extract the unit historical transaction data corresponding to each abnormal dimension from the historical transaction data; Determine the detection rules corresponding to each abnormal dimension respectively; Detect the unit historical transaction data corresponding to each abnormal dimension respectively according to the detection rules corresponding to each abnormal dimension to obtain the abnormal transaction data images under each abnormal dimension.

7. The method according to claim 6, wherein The unit historical transaction data corresponding to the target abnormal dimension includes transaction records, and the number of transaction records is multiple; The process of detecting the unit historical transaction data corresponding to the target abnormal dimension according to the detection rule corresponding to the target abnormal dimension to obtain the abnormal transaction data image under the target abnormal dimension includes: Detecting multiple transaction records according to the detection rule corresponding to the target abnormal dimension to obtain the abnormal transaction records among the multiple transaction records; Generating the abnormal transaction data image under the target abnormal dimension, where the abnormal transaction data image under the target abnormal dimension includes the abnormal transaction records.

8. The method according to claim 7, wherein The generating of the abnormal transaction data image under the target abnormal dimension includes: Generating a statistical analysis result according to the abnormal transaction records, the multiple transaction records, and the detection rule corresponding to the target abnormal dimension; Generating the abnormal transaction data image under the target abnormal dimension, where the abnormal transaction data image under the target abnormal dimension includes the abnormal transaction records and the statistical analysis result.

9. The method according to claim 8, wherein It further includes: Determining the risk prediction label of the abnormal transaction records according to the statistical analysis result; Setting an association relationship between the risk prediction label of the abnormal transaction records and the abnormal transaction data image under the target abnormal dimension.

10. The method according to claim 6, characterized in that, The unit historical transaction data corresponding to the target abnormal dimension includes transaction texts, and the number of transaction texts is multiple; The process of detecting the unit historical transaction data corresponding to the target abnormal dimension according to the detection rule corresponding to the target abnormal dimension to obtain the abnormal transaction data image under the target abnormal dimension includes: Detecting multiple transaction texts according to the detection rule corresponding to the target abnormal dimension to obtain the abnormal transaction texts among the multiple transaction texts; Generating the abnormal transaction data image under the target abnormal dimension, where the abnormal transaction data image under the target abnormal dimension includes the abnormal transaction texts.

11. The method according to claim 10, characterized in that, The detecting of the multiple transaction texts according to the detection rule corresponding to the target abnormal dimension to obtain the abnormal transaction texts among the multiple transaction texts includes: Matching the abnormal keyword library with each transaction text according to the detection rule corresponding to the target abnormal dimension to obtain abnormal keywords; Regarding the transaction text where the abnormal keyword is located as the abnormal transaction text.

12. The method according to claim 11, wherein The generating of the abnormal transaction data image under the target abnormal dimension includes: Obtaining the risk prediction label corresponding to the abnormal keyword; Generating the abnormal transaction data image under the target abnormal dimension, where the abnormal transaction data image under the target abnormal dimension includes the abnormal transaction texts and the risk prediction label corresponding to the abnormal keyword, and the abnormal keyword is marked in the abnormal transaction data image under the target abnormal dimension according to a preset marking method.

13. An image processing apparatus, characterized in that, It includes: A first display module for displaying a transaction monitoring interface, where the transaction monitoring interface includes historical transaction data and an image display control; A second display module, configured to display an abnormal result interface when the image display control is selected; the abnormal result interface includes one or more abnormal dimensions, and one or more abnormal transaction data images are included under each abnormal dimension. The abnormal transaction data images under each abnormal dimension are obtained by detecting the historical transaction data using a detection rule corresponding to each abnormal dimension. A third display module, configured to display a preview interface of the target abnormal dimension and display one or more abnormal transaction data images under the target abnormal dimension in the preview interface of the target abnormal dimension when there is a preview operation for the target abnormal dimension, where the target abnormal dimension is one of the one or more abnormal dimensions.

14. A computer device, characterized in that, It includes a memory and a processor. The memory stores a computer program. When the computer program is executed by the processor, the processor is caused to execute the steps of the method according to any one of claims 1-12.

15. A computer storage medium, characterized in that, The computer storage medium stores a computer program, and the computer program includes program instructions. When the program instructions are executed by a processor, a computer device having the processor is caused to execute the method according to any one of claims 1-12.

Images