Abnormal transaction identification method and device, storage medium and electronic equipment

Abstract

The invention discloses an abnormal transaction identification method and device, a storage medium and electronic equipment. The method relates to the field of financial science and technology, and comprises the following steps: obtaining multi-modal data corresponding to a target transaction, and determining a target transaction scene corresponding to the target transaction according to the multi-modal data; under the condition that the target transaction scene is the first transaction scene, performing feature extraction processing on operation behavior data in the image data, the voice data and the environment data to obtain image features, voice features and operation behavior features; performing weighted fusion processing according to the image features, the voice features and the operation behavior features to obtain fusion data; and judging whether the target transaction is abnormal or not according to the fusion data and a preset threshold value corresponding to the target transaction scene, and obtaining a judgment result. Through application of the method and the device, the problem that the recognition accuracy is relatively low in a strong light environment and a face shielding scene due to the fact that abnormal transaction recognition depends on a single visual mode in related technologies is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of financial technology, in particular, to an abnormal transaction identification method and device, a storage medium and an electronic device. BACKGROUND

[0002] In the current Automated Teller Machine (ATM) operation, in order to ensure transaction security, the abnormal transaction identification technology has become an indispensable part. In related technologies, the ATM abnormal transaction identification system mainly relies on a single visual modality, and identifies abnormal situations by analyzing the facial expressions of users. However, in a strong light environment (such as sunlight directly shining on the ATM screen), the overexposure of the face image will cause feature loss, resulting in low accuracy of abnormal identification. In addition, in the occlusion scene (such as the user wearing a mask and operating with his head down), the key area of the face is blocked, and the misjudgment rate is high.

[0003] In related technologies, the single visual modality is used for abnormal transaction identification, and in strong light environment and face occlusion scene, the identification accuracy is low, and there is no effective solution at present. SUMMARY

[0004] The main purpose of the present application is to provide an abnormal transaction identification method and device, a storage medium and an electronic device, to solve the problem of low identification accuracy in related technologies that rely on a single visual modality for abnormal transaction identification in strong light environment and face occlusion scene.

[0005] In order to achieve the above purpose, according to one aspect of the present application, an abnormal transaction identification method is provided, applied to an Automated Teller Machine, the method comprising: acquiring multi-modal data corresponding to a target transaction, and determining a target transaction scene corresponding to the target transaction according to the multi-modal data, wherein the multi-modal data includes image data, voice data and environment data; in the case that the target transaction scene is a first transaction scene, performing feature extraction processing on operation behavior data in the image data, the voice data and the environment data respectively to obtain image features, voice features and operation behavior features, wherein the first transaction scene represents a scene in which the light intensity is greater than a light threshold and the face occlusion area of a transaction object initiating the target transaction is greater than an occlusion threshold; performing weighted fusion processing on the image features, the voice features and the operation behavior features to obtain fusion data, wherein the fusion data is used to represent the abnormality degree of the target transaction; determining whether the target transaction is abnormal according to the fusion data and a preset threshold corresponding to the target transaction scene to obtain a judgment result.

[0006] Further, in the case that the target transaction scenario is the first transaction scenario, the feature extraction processing is respectively performed on the image data, the voice data and the operation behavior data in the environment data to obtain image features, voice features and operation behavior features, including: generating a fusion image according to the visible light image and the infrared image in the image data, performing feature extraction processing on the fusion image to obtain the image features; performing noise reduction processing on the voice data to obtain noise-reduced voice data, performing feature extraction processing on the noise-reduced voice data to obtain the voice features; performing statistical analysis on the operation behavior data to obtain statistical data, performing feature extraction processing on the statistical data to obtain the operation behavior features.

[0007] Further, the generating of the fusion image according to the visible light image and the infrared image in the image data includes: performing light correction processing on the visible light image to obtain a corrected visible light image; performing image enhancement processing on the corrected visible light image by using a preset model to obtain an enhanced visible light image; and performing weighted fusion processing on the enhanced visible light image and the infrared image to obtain the fusion image.

[0008] Further, the weighted fusion processing according to the image features, the voice features and the operation behavior features to obtain the fusion data includes: respectively calculating the credibility of the image features, the credibility of the voice features and the credibility of the operation behavior features; determining the weight of the image features, the weight of the voice features and the weight of the operation behavior features according to the credibility of the image features, the credibility of the voice features and the credibility of the operation behavior features; and performing weighted fusion processing according to the image features, the weight of the image features, the voice features, the weight of the voice features, the operation behavior features and the weight of the operation behavior features to obtain the fusion data.

[0009] Further, the judging of whether the target transaction is abnormal according to the fusion data and the preset threshold corresponding to the target transaction scenario to obtain a judgment result includes: comparing the fusion data with the preset threshold; if the fusion data is greater than the preset threshold, taking the target transaction as abnormal as the judgment result; and if the fusion data is less than or equal to the preset threshold, taking the target transaction as normal as the judgment result.

[0010] Further, after taking the target transaction as abnormal as the judgment result, the method further includes: calculating a difference between the fusion data and the preset threshold, and comparing the difference with a warning threshold; if the difference is greater than the warning threshold, performing a first warning strategy; and if the difference is less than or equal to the warning threshold, performing a second warning strategy, wherein the first warning strategy and the second warning strategy correspond to different warning processing procedures.

[0011] Further, before acquiring the multi-modal data corresponding to the target transaction, the method further includes: acquiring images through a camera and an infrared camera of the automatic teller machine to obtain a visible light image and an infrared image, generating image data according to the visible light image and the infrared image; acquiring voice through a voice acquisition module of the automatic teller machine to obtain voice data; acquiring illumination through an ambient light sensor of the automatic teller machine to obtain an illumination intensity, and acquiring operation behavior through an operation behavior acquisition module of the automatic teller machine to obtain operation behavior data, generating environment data according to the illumination intensity and the operation behavior data.

[0012] To achieve the above object, according to another aspect of the present application, an abnormal transaction identification device is provided, which is applied to an automatic teller machine, and the device comprises: an acquisition unit configured to acquire multi-modal data corresponding to a target transaction, and determine a target transaction scene corresponding to the target transaction according to the multi-modal data, wherein the multi-modal data comprises image data, voice data and environment data; a first processing unit configured to, in the case that the target transaction scene is a first transaction scene, perform feature extraction processing on the image data, the voice data and operation behavior data in the environment data respectively to obtain image features, voice features and operation behavior features, wherein the first transaction scene represents a scene in which an illumination intensity is greater than an illumination threshold and a face occlusion area of a transaction object initiating the target transaction is greater than an occlusion threshold; a second processing unit configured to perform weighted fusion processing on the image features, the voice features and the operation behavior features to obtain fusion data, wherein the fusion data is used to represent an abnormality degree of the target transaction; and a third processing unit configured to determine whether the target transaction is abnormal according to the fusion data and a preset threshold corresponding to the target transaction scene to obtain a determination result.

[0013] Further, the first processing unit comprises: a first processing sub-unit configured to generate a fusion image according to a visible light image and an infrared image in the image data, perform feature extraction processing on the fusion image to obtain image features; a second processing sub-unit configured to perform noise reduction processing on the voice data to obtain noise-reduced voice data, perform feature extraction processing on the noise-reduced voice data to obtain voice features; and a third processing sub-unit configured to perform statistical analysis on the operation behavior data to obtain statistical data, and perform feature extraction processing on the statistical data to obtain operation behavior features.

[0014] Further, the first processing sub-unit comprises: a first processing module configured to perform light correction processing on the visible light image to obtain a corrected visible light image; a second processing module configured to perform image enhancement processing on the corrected visible light image using a preset model to obtain an enhanced visible light image; and a third processing module configured to perform weighted fusion processing on the enhanced visible light image and the infrared image to obtain the fusion image.

[0015] Further, the second processing unit includes: a fourth processing subunit, used to calculate the credibility of image features, the credibility of speech features, and the credibility of operation behavior features respectively; a fifth processing subunit, used to determine the weights of image features, speech features, and operation behavior features based on the credibility of image features, the credibility of speech features, and the credibility of operation behavior features; and a sixth processing subunit, used to perform weighted fusion processing based on image features, the weights of image features, speech features, the weights of speech features, operation behavior features, and the weights of operation behavior features to obtain fused data.

[0016] Furthermore, the third processing unit includes: a seventh processing subunit for comparing the fused data with a preset threshold; a first determining subunit for determining that if the fused data is greater than the preset threshold, the target transaction has an anomaly as the judgment result; and a second determining subunit for determining that if the fused data is less than or equal to the preset threshold, the target transaction does not have an anomaly as the judgment result.

[0017] Furthermore, the device also includes: a calculation unit, used to calculate the difference between the fused data and a preset threshold after determining that the target transaction is abnormal, and to compare the difference with a warning threshold; a first execution unit, used to execute a first warning strategy if the difference is greater than the warning threshold; and a second execution unit, used to execute a second warning strategy if the difference is less than or equal to the warning threshold, wherein the warning processing procedures corresponding to the first warning strategy and the second warning strategy are different.

[0018] Furthermore, the device also includes: a first acquisition unit, used to acquire images through the ATM's camera and infrared camera before acquiring the multimodal data corresponding to the target transaction, to obtain visible light images and infrared images, and to generate image data based on the visible light images and infrared images; a second acquisition unit, used to acquire voice data through the ATM's voice acquisition module; and a third acquisition unit, used to acquire light intensity through the ATM's ambient light sensor, and to acquire operation behavior data through the ATM's operation behavior acquisition module, and to generate environmental data based on the light intensity and operation behavior data.

[0019] According to another aspect of the present invention, an electronic device is also provided, comprising: a memory storing an executable program; and a processor for running the program, wherein the program executes the abnormal transaction identification method of any one of the above-mentioned methods during runtime.

[0020] According to another aspect of the present invention, a computer-readable storage medium is also provided, wherein a program is stored in the storage medium, and a method for identifying abnormal transactions that controls the device where the storage medium is located to execute any of the above-mentioned transactions is provided when the program is running.

[0021] In this embodiment, the following steps are employed: First, multimodal data corresponding to the target transaction is acquired, and the target transaction scenario corresponding to the target transaction is determined based on the multimodal data. The multimodal data includes image data, voice data, and environmental data. Second, when the target transaction scenario is a first transaction scenario, feature extraction processing is performed on the operational behavior data in the image data, voice data, and environmental data to obtain image features, voice features, and operational behavior features. The first transaction scenario represents a scenario where the light intensity is greater than a light threshold and the facial occlusion area of ​​the transaction object initiating the target transaction is greater than an occlusion threshold. Third, weighted fusion processing is performed on the image features, voice features, and operational behavior features to obtain fused data. The fused data is used to characterize the degree of abnormality of the target transaction. Finally, based on the fused data and a preset threshold corresponding to the target transaction scenario, it is determined whether the target transaction is abnormal, and a judgment result is obtained. This solves the technical problem in related technologies where relying on a single visual modality for abnormal transaction identification results in low recognition accuracy in strong light environments and facial occlusion scenarios. In this solution, multimodal fusion of vision, voice, and behavior can effectively cope with strong light and occlusion scenarios, ensuring accurate judgment of abnormal transaction situations under these conditions, improving the accuracy of abnormal transaction identification, enhancing anti-interference capabilities, and reducing false alarm rate. Attached Figure Description

[0022] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0023] Figure 1 A hardware structure block diagram of a computer terminal for implementing a method for identifying abnormal transactions is shown.

[0024] Figure 2 This is a flowchart of an abnormal transaction identification method provided according to an embodiment of this application;

[0025] Figure 3 This is a schematic diagram of the architecture of an abnormal transaction identification system provided according to an embodiment of this application;

[0026] Figure 4 This is a schematic diagram of an abnormal transaction identification device provided according to an embodiment of this application;

[0027] Figure 5 This is a structural block diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0028] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0029] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0030] It should be noted that the information collected in this application (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for display, data used for analysis, etc.) are information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of this data all comply with relevant laws, regulations, and standards, necessary confidentiality measures have been taken, and they do not violate public order and good morals. Corresponding access points are provided for users to choose to authorize or refuse. For example, interfaces are set up between this system and relevant users or organizations, providing users with corresponding access points to choose to agree to or refuse automated decision-making results; if the user chooses to refuse, the process proceeds to the expert decision-making stage.

[0031] Example 1

[0032] According to an embodiment of this application, a method embodiment for identifying abnormal transactions is also provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0033] The method embodiment provided in Embodiment 1 of this application can be executed on a mobile terminal, computer terminal, or similar computing device. Figure 1A hardware block diagram of a computer terminal (or mobile device) for implementing a method for identifying abnormal transactions is shown. Figure 1 As shown, the computer terminal 10 (or mobile device) may include one or more processors 102 (shown as 102a, 102b, ..., 102n in the figure) 102 (processor 102 may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of a BUS bus), a network interface, a power supply, and / or a camera. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, computer terminal 10 may also include... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0034] It should be noted that the aforementioned one or more processors 102 and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be integrated, in whole or in part, into any other element within the computer terminal 10 (or mobile device). As involved in the embodiments of this application, the data processing circuits serve as a processor control mechanism (e.g., selection of a variable resistor termination path connected to an interface).

[0035] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the abnormal transaction identification method in this embodiment. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby realizing the above-mentioned abnormal transaction identification method. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0036] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the computer terminal 10. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0037] The display may be a touchscreen liquid crystal display (LCD) that allows the user to interact with the user interface of the computer terminal 10 (or mobile device).

[0038] Under the aforementioned operating environment, this application provides the following: Figure 2 The method for identifying abnormal transactions is shown. Figure 2 This is a flowchart of an abnormal transaction identification method according to Embodiment 1 of this application. The abnormal transaction identification method includes:

[0039] Step S201: Obtain multimodal data corresponding to the target transaction, and determine the target transaction scenario corresponding to the target transaction based on the multimodal data. The multimodal data includes image data, voice data, and environmental data.

[0040] Step S202: When the target transaction scenario is the first transaction scenario, feature extraction processing is performed on the operation behavior data in the image data, voice data and environmental data respectively to obtain image features, voice features and operation behavior features. The first transaction scenario represents a scenario where the light intensity is greater than the light threshold and the face occlusion area of ​​the transaction object initiating the target transaction is greater than the occlusion threshold.

[0041] Step S203: Perform weighted fusion processing based on image features, voice features, and operation behavior features to obtain fused data, wherein the fused data is used to characterize the degree of anomaly of the target transaction;

[0042] Step S204: Based on the fused data and the preset threshold corresponding to the target transaction scenario, determine whether there is any abnormality in the target transaction and obtain the judgment result.

[0043] Optionally, the ATM abnormal transaction identification system acquires multimodal data corresponding to the current transaction (i.e., the target transaction), such as image data (visible light and infrared images), voice data, and environmental data (light intensity and ATM operation behavior). Then, the target transaction scenario corresponding to the current transaction is determined; for example, the light intensity is used to determine if it is a strong light environment, and the image data is used to determine if it is a scene with facial occlusion.

[0044] Optionally, the first transaction scenario is a strong light environment and a face occlusion scenario. For example, when the light intensity is greater than 5000 lux (light threshold), it is considered a strong light environment; when the face occlusion area is greater than 30% (occlusion threshold), it is considered a face occlusion scenario. In strong light environment and face occlusion scenario, the system performs feature extraction processing on the image data, voice data, and operation behavior data in the environmental data respectively to obtain image features, voice features, and operation behavior features. Then, based on the image features, voice features, and operation behavior features, a weighted fusion processing is performed to obtain fusion outliers (i.e., fused data).

[0045] Optionally, based on the fused outliers and a preset threshold corresponding to the target transaction scenario, it can be determined whether the target transaction is abnormal, and a judgment result can be obtained. For example, in a normal transaction scenario, the preset threshold is 0.8, and in a strong light environment or a scenario with facial occlusion, the preset threshold is 0.6. By comparing the fused outliers with the preset threshold, it can be determined whether the current transaction is abnormal, thus realizing the identification of abnormal transactions.

[0046] In summary, multimodal fusion of vision, voice, and behavior can effectively address scenarios with strong light and occlusion, ensuring accurate identification of abnormal transactions under these conditions. This improves the accuracy of abnormal transaction identification, enhances anti-interference capabilities, and reduces the false alarm rate.

[0047] Optionally, in the abnormal transaction identification method provided in this application embodiment, when the target transaction scenario is the first transaction scenario, feature extraction processing is performed on the operation behavior data in the image data, voice data, and environmental data respectively to obtain image features, voice features, and operation behavior features, including: generating a fused image based on the visible light image and infrared image in the image data, performing feature extraction processing on the fused image to obtain image features; performing noise reduction processing on the voice data to obtain denoised voice data, performing feature extraction processing on the denoised voice data to obtain voice features; performing statistical analysis on the operation behavior data to obtain statistical data, and performing feature extraction processing on the statistical data to obtain operation behavior features.

[0048] In an optional embodiment, the feature extraction submodule performs a strong light preprocessing process and an occlusion preprocessing process. It generates a fused image based on the visible light image and infrared image in the image data, performs feature extraction processing on the fused image to obtain image features. For example, it uses a deep learning model to identify the occluded area of ​​the face of the transaction object, calculates the occlusion ratio, and performs feature enhancement on the unoccluded facial area to extract visual features such as pupil dilation rate and frowning degree. At the same time, it performs weighted fusion of infrared thermal imaging and visible light image to extract features such as facial contour and blood vessel distribution that are not affected by occlusion.

[0049] Optionally, the speech data is denoised to obtain denoised speech data, and then feature extraction is performed on the denoised speech data to obtain speech features. For example, a denoising algorithm is used to improve the quality of the speech signal, and then a recognition model trained by a long short-term memory neural network is used to extract temporal emotional features such as fundamental frequency, short-time energy, and Mel-frequency cepstral coefficients from the speech signal, thereby obtaining probability values ​​containing emotions such as anger, fear, and anxiety.

[0050] Optionally, statistical analysis can be performed on the operational behavior data to obtain statistical data, and feature extraction processing can be performed on the statistical data to obtain operational behavior features. For example, analyzing the operations of a transaction object on an ATM, such as key press intervals, number of amount modifications, and number of consecutive withdrawal limit entries, can quantify and generate operational behavior feature values ​​that reflect the degree of operational abnormality.

[0051] By employing a multimodal data fusion strategy for feature extraction, the accuracy and robustness of the identification were enhanced, and the false alarm rate was reduced.

[0052] Optionally, in the abnormal transaction identification method provided in this application embodiment, generating a fused image based on the visible light image and infrared image in the image data includes: performing light correction processing on the visible light image to obtain a corrected visible light image; performing image enhancement processing on the corrected visible light image using a preset model to obtain an enhanced visible light image; and performing weighted fusion processing on the enhanced visible light image and the infrared image to obtain a fused image.

[0053] In an optional embodiment, during the generation of the fused image, the visible light image is first subjected to light correction processing to obtain a corrected visible light image. Then, a preset deep learning model is used to perform image enhancement processing on the corrected visible light image to obtain an enhanced visible light image. Finally, the enhanced visible light image and the infrared image are subjected to weighted fusion processing to obtain the fused image.

[0054] By generating fused images, details in key facial regions are enhanced, thereby improving the accuracy of identifying abnormal transactions.

[0055] Optionally, in the abnormal transaction identification method provided in this application embodiment, the weighted fusion processing based on image features, voice features, and operation behavior features to obtain fused data includes: calculating the credibility of image features, voice features, and operation behavior features respectively; determining the weights of image features, voice features, and operation behavior features based on the credibility of image features, voice features, and operation behavior features; and performing weighted fusion processing based on image features, image feature weights, voice features, voice feature weights, operation behavior features, and operation behavior feature weights to obtain fused data.

[0056] In an optional embodiment, the credibility of image features, speech features, and operational behavior features are calculated separately. For example, visual credibility = 1 - occlusion ratio, and speech credibility = the sentiment probability output by the long short-term memory neural network model. Based on the credibility of each modality, the weight ratios of visual, speech, and behavioral features in the fusion calculation are adjusted in real time. For example, when the occlusion ratio is >50%, the speech feature weight is increased to 0.6, and the visual feature weight is decreased to 0.2 to ensure the reliability of the fusion result. Then, a weighted fusion process is performed based on image features, image feature weights, speech features, speech feature weights, operational behavior features, and operational behavior feature weights to obtain fusion outliers. Optionally, the formula for calculating fusion outliers is as follows:

[0057]

[0058] in, The weights of image features, This refers to facial outliers (i.e., visual credibility). The weights of the speech features, This is an abnormal voice value. The weights of the operational behavior features, This is an abnormal value in the operation behavior.

[0059] By identifying fused outliers, the degree of transaction anomalies of current users can be more comprehensively reflected, thereby effectively avoiding misjudgments caused by limitations of single-modal data and enhancing the system's environmental adaptability and anti-interference capabilities.

[0060] Optionally, in the abnormal transaction identification method provided in this application embodiment, the determination of whether the target transaction is abnormal based on the fused data and the preset threshold corresponding to the target transaction scenario includes: comparing the fused data and the preset threshold; if the fused data is greater than the preset threshold, then the target transaction is considered to be abnormal as the determination result; if the fused data is less than or equal to the preset threshold, then the target transaction is considered to be not abnormal as the determination result.

[0061] In an optional embodiment, the fused outlier is compared with a preset threshold (e.g., 0.8 in a normal transaction scenario, and 0.6 in a strong light environment and a face occlusion scenario). If the fused data is greater than the preset threshold, the target transaction is considered to have an anomaly. If the fused data is less than or equal to the preset threshold, the target transaction is considered to have no anomaly.

[0062] Optionally, in the abnormal transaction identification method provided in this application embodiment, after determining that the target transaction is abnormal, the method further includes: calculating the difference between the fused data and a preset threshold, and comparing the difference with a warning threshold; if the difference is greater than the warning threshold, then executing a first warning strategy; if the difference is less than or equal to the warning threshold, then executing a second warning strategy, wherein the warning processing procedures corresponding to the first warning strategy and the second warning strategy are different.

[0063] In an optional embodiment, the intervention level is set based on the difference between the fused anomaly and the dynamic threshold. For example, when the difference between the fused anomaly and the dynamic threshold is greater than 0.3 (the warning threshold), a level one warning (i.e., the first warning strategy) is triggered, such as triggering forced transaction termination, audible and visual alarms, and notification to the remote monitoring center. When the difference is less than or equal to the warning threshold, a level two warning is triggered, such as triggering transaction delays and remote auditing.

[0064] The tiered early warning mechanism enables a more accurate matching of anomaly severity and response measures, thereby enhancing system security.

[0065] Optionally, in the abnormal transaction identification method provided in this application embodiment, before obtaining the multimodal data corresponding to the target transaction, the method further includes: acquiring images through the camera and infrared camera of the ATM to obtain visible light images and infrared images, and generating image data based on the visible light images and infrared images; acquiring voice data through the voice acquisition module of the ATM; acquiring light intensity through the ambient light sensor of the ATM, and acquiring operation behavior data through the operation behavior acquisition module of the ATM, and generating environmental data based on the light intensity and operation behavior data.

[0066] In an optional embodiment, multi-source data is first collected synchronously. This involves acquiring images using the ATM's camera and infrared camera to obtain visible light and infrared images, and then generating image data based on these images. Voice data is then acquired using the ATM's voice acquisition module (microphone array). Light intensity is obtained using the ATM's ambient light sensor, and operational behavior data is acquired using the ATM's operation behavior acquisition module (operation behavior collector). Environmental data is then generated based on the light intensity and operation behavior data.

[0067] In an optional embodiment, Figure 3 This is a schematic diagram of the architecture of an abnormal transaction identification system provided in the embodiments of this application, such as... Figure 3As shown, the system includes a multi-source perception module, a preprocessing module, a multi-modal feature fusion module, a dynamic threshold decision module, and an execution module. The multi-source perception module comprises a visual perception submodule (main camera and infrared camera), a voice perception submodule (microphone array signal), and an environment and behavior perception submodule (ambient light sensor and ATM operation behavior collector). First, multi-source data is acquired synchronously: visible light and infrared images are acquired through the main camera and infrared camera; voice data is acquired through the microphone; and environmental data such as light intensity and ATM operation behavior are acquired through the ambient light sensor and ATM operation behavior collector.

[0068] Optionally, the preprocessing module includes an image preprocessing submodule and a speech preprocessing submodule. Based on the collected light intensity, it determines whether the scene is in strong light. If so, a strong light preprocessing process is performed. The image preprocessing submodule uses the Pettine algorithm to suppress strong light, divides the face into three regions: forehead, eyes, and cheeks, calculates the optimal exposure parameters for each region to avoid local overexposure, and weighted fuses the visible light image and infrared image to obtain an enhanced facial image. If not, or after obtaining the enhanced facial image, it determines whether the scene is occluded. If so, an occlusion preprocessing process is performed. The image preprocessing submodule generates an occlusion mask using a convolutional neural network, enhances feature extraction of unoccluded areas such as eyes and eyebrows using an attention mechanism, and captures facial temperature distribution using an infrared camera. Infrared-assisted recognition compensates for the lack of visible light features to obtain the final visible area features. If not, a conventional processing flow is executed, directly extracting complete facial features. The speech preprocessing submodule performs noise reduction processing on the collected speech data to obtain processed speech data.

[0069] Optionally, the multimodal feature fusion module includes a feature extraction submodule and a dynamic weight allocation submodule. The feature extraction submodule extracts features from the preprocessed data to obtain visual features, speech features, and behavioral features. The dynamic weight allocation submodule calculates the credibility of each modality, dynamically allocates weights based on the credibility, and calculates the fusion anomaly using the fusion anomaly calculation formula.

[0070] Optionally, the dynamic threshold decision module includes an environmental credibility assessment submodule, a threshold adjustment submodule, and an early warning signal generation submodule. The environmental credibility assessment submodule performs illumination and occlusion assessments. If it is a scene with strong light and occlusion, the threshold adjustment submodule determines the current threshold based on the current scene information; for example, in a scene with strong light and occlusion, the threshold is 0.6. The early warning signal generation submodule generates early warning signals. When the fused outlier is greater than the threshold, it calculates the difference between the fused outlier and the threshold. For example, if the difference is greater than the early warning threshold, a level one early warning signal is output.

[0071] Optionally, the execution module includes a transaction intervention submodule and an alarm submodule. The transaction intervention submodule is used for transaction delay and remote auditing, and the alarm submodule is used for audible and visual alarms and forced termination.

[0072] The abnormal transaction identification method provided in this application includes the following steps: acquiring multimodal data corresponding to the target transaction, and determining the target transaction scenario based on the multimodal data, wherein the multimodal data includes image data, voice data, and environmental data; when the target transaction scenario is a first transaction scenario, performing feature extraction processing on the operational behavior data in the image data, voice data, and environmental data respectively to obtain image features, voice features, and operational behavior features, wherein the first transaction scenario represents a scenario where the light intensity is greater than a light threshold and the facial occlusion area of ​​the transaction object initiating the target transaction is greater than an occlusion threshold; performing weighted fusion processing on the image features, voice features, and operational behavior features to obtain fused data, wherein the fused data is used to characterize the degree of abnormality of the target transaction; and determining whether the target transaction is abnormal based on the fused data and a preset threshold corresponding to the target transaction scenario to obtain a judgment result. This solves the technical problem in related technologies that rely on a single visual modality for abnormal transaction identification, resulting in low recognition accuracy in strong light environments and facial occlusion scenarios. In this solution, multimodal fusion of vision, voice, and behavior can effectively cope with strong light and occlusion scenarios, ensuring accurate judgment of abnormal transaction situations under these conditions, improving the accuracy of abnormal transaction identification, enhancing anti-interference capabilities, and reducing false alarm rate.

[0073] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0074] Example 2

[0075] This application also provides an abnormal transaction identification device. It should be noted that the abnormal transaction identification device of this application can be used to execute the abnormal transaction identification method provided in this application. The abnormal transaction identification device provided in this application will be described below.

[0076] According to an embodiment of this application, an abnormal transaction identification device for implementing the above-described abnormal transaction identification method is also provided, such as... Figure 4 As shown, the device includes: an acquisition unit 401, a first processing unit 402, a second processing unit 403, and a third processing unit 404.

[0077] The acquisition unit 401 is used to acquire multimodal data corresponding to the target transaction and determine the target transaction scenario corresponding to the target transaction based on the multimodal data. The multimodal data includes image data, voice data and environmental data.

[0078] The first processing unit 402 is used to perform feature extraction processing on the operation behavior data in the image data, voice data and environmental data respectively when the target transaction scenario is the first transaction scenario, to obtain image features, voice features and operation behavior features. The first transaction scenario represents a scenario where the light intensity is greater than the light threshold and the face occlusion area of ​​the transaction object initiating the target transaction is greater than the occlusion threshold.

[0079] The second processing unit 403 is used to perform weighted fusion processing based on image features, voice features and operation behavior features to obtain fused data, wherein the fused data is used to characterize the degree of abnormality of the target transaction.

[0080] The third processing unit 404 is used to determine whether there is an anomaly in the target transaction based on the fused data and the preset threshold corresponding to the target transaction scenario, and obtain the judgment result.

[0081] The abnormal transaction identification device provided in this application embodiment acquires multimodal data corresponding to the target transaction through an acquisition unit 401, and determines the target transaction scenario corresponding to the target transaction based on the multimodal data. The multimodal data includes image data, voice data, and environmental data. When the target transaction scenario is a first transaction scenario, the first processing unit 402 performs feature extraction processing on the operation behavior data in the image data, voice data, and environmental data respectively to obtain image features, voice features, and operation behavior features. The first transaction scenario represents a scenario where the light intensity is greater than the light threshold and the facial occlusion area of ​​the transaction object initiating the target transaction is greater than the occlusion threshold. The second processing unit 403 performs weighted fusion processing on the image features, voice features, and operation behavior features to obtain fused data. The fused data is used to characterize the degree of abnormality of the target transaction. The third processing unit 404 determines whether the target transaction is abnormal based on the fused data and the preset threshold corresponding to the target transaction scenario, and obtains a judgment result. By integrating visual, voice, and behavioral multimodal data, it can effectively cope with strong light and occlusion scenarios, ensuring accurate judgment of abnormal transactions under these conditions. This improves the accuracy of abnormal transaction identification, enhances anti-interference capabilities, and reduces false alarm rates.

[0082] Optionally, in the abnormal transaction identification device provided in this application embodiment, the first processing unit includes: a first processing subunit, used to generate a fused image based on the visible light image and infrared image in the image data, and to perform feature extraction processing on the fused image to obtain image features; a second processing subunit, used to perform noise reduction processing on the speech data to obtain denoised speech data, and to perform feature extraction processing on the denoised speech data to obtain speech features; and a third processing subunit, used to perform statistical analysis on the operation behavior data to obtain statistical data, and to perform feature extraction processing on the statistical data to obtain operation behavior features.

[0083] Optionally, in the abnormal transaction identification device provided in this application embodiment, the first processing subunit includes: a first processing module, used to perform light correction processing on a visible light image to obtain a corrected visible light image; a second processing module, used to perform image enhancement processing on the corrected visible light image using a preset model to obtain an enhanced visible light image; and a third processing module, used to perform weighted fusion processing on the enhanced visible light image and the infrared image to obtain a fused image.

[0084] Optionally, in the abnormal transaction identification device provided in this application embodiment, the second processing unit includes: a fourth processing subunit, used to calculate the credibility of image features, the credibility of voice features, and the credibility of operation behavior features respectively; a fifth processing subunit, used to determine the weights of image features, voice features, and operation behavior features based on the credibility of image features, the credibility of voice features, and the credibility of operation behavior features; and a sixth processing subunit, used to perform weighted fusion processing based on image features, the weights of image features, voice features, the weights of voice features, operation behavior features, and the weights of operation behavior features to obtain fused data.

[0085] Optionally, in the abnormal transaction identification device provided in the embodiments of this application, the third processing unit includes: a seventh processing subunit, used to compare the fused data with a preset threshold; a first determining subunit, used to determine that the target transaction is abnormal if the fused data is greater than the preset threshold; and a second determining subunit, used to determine that the target transaction is not abnormal if the fused data is less than or equal to the preset threshold.

[0086] Optionally, in the abnormal transaction identification device provided in the embodiments of this application, the device further includes: a calculation unit, used to calculate the difference between the fused data and a preset threshold after determining that the target transaction is abnormal, and to compare the difference with a warning threshold; a first execution unit, used to execute a first warning strategy if the difference is greater than the warning threshold; and a second execution unit, used to execute a second warning strategy if the difference is less than or equal to the warning threshold, wherein the warning processing procedures corresponding to the first warning strategy and the second warning strategy are different.

[0087] Optionally, in the abnormal transaction identification device provided in this application embodiment, the device further includes: a first acquisition unit, used to acquire images through the camera and infrared camera of the ATM before acquiring the multimodal data corresponding to the target transaction, to obtain visible light images and infrared images, and to generate image data based on the visible light images and infrared images; a second acquisition unit, used to acquire voice data through the voice acquisition module of the ATM; and a third acquisition unit, used to acquire light intensity through the ambient light sensor of the ATM, and to acquire operation behavior data through the operation behavior acquisition module of the ATM, and to generate environmental data based on the light intensity and operation behavior data.

[0088] It should be noted that the acquisition unit 401, the first processing unit 402, the second processing unit 403, and the third processing unit 404 mentioned above correspond to steps S201 to S204 in Embodiment 1. The four units and the corresponding steps implement the same instances and application scenarios, but are not limited to the content disclosed in Embodiment 1. It should be noted that the above modules or units can be hardware or software components stored in memory (e.g., memory 104) and processed by one or more processors (e.g., processors 102a, 102b, ..., 102n). The above units can also be part of a device and run in the computer terminal 10 provided in Embodiment 1.

[0089] Example 3

[0090] Embodiments of this application may provide an electronic device. Figure 5 This is a structural block diagram of an electronic device according to an embodiment of this application. Figure 5 As shown, the electronic device may include: one or more ( Figure 5 (Only one is shown) processor 502, memory 504, memory controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module and display.

[0091] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the methods and apparatus in the embodiments of this application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby implementing the above-described methods. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0092] The processor can invoke information and applications stored in the memory via a transmission device to execute the following steps: acquiring multimodal data corresponding to the target transaction, and determining the target transaction scenario based on the multimodal data, wherein the multimodal data includes image data, voice data, and environmental data; when the target transaction scenario is a first transaction scenario, performing feature extraction processing on the operational behavior data in the image data, voice data, and environmental data respectively to obtain image features, voice features, and operational behavior features, wherein the first transaction scenario represents a scenario where the light intensity is greater than the light threshold and the facial occlusion area of ​​the transaction object initiating the target transaction is greater than the occlusion threshold; performing weighted fusion processing on the image features, voice features, and operational behavior features to obtain fused data, wherein the fused data is used to characterize the degree of abnormality of the target transaction; and judging whether the target transaction is abnormal based on the fused data and the preset threshold corresponding to the target transaction scenario to obtain a judgment result.

[0093] The processor can call the information and application program stored in the memory through the transmission device to perform the following steps: generate a fused image based on the visible light image and infrared image in the image data, perform feature extraction processing on the fused image to obtain image features; perform noise reduction processing on the speech data to obtain denoised speech data, perform feature extraction processing on the denoised speech data to obtain speech features; perform statistical analysis on the operation behavior data to obtain statistical data, and perform feature extraction processing on the statistical data to obtain operation behavior features.

[0094] The processor can call the information and application program stored in the memory through the transmission device to perform the following steps: perform light correction processing on the visible light image to obtain a corrected visible light image; perform image enhancement processing on the corrected visible light image using a preset model to obtain an enhanced visible light image; and perform weighted fusion processing on the enhanced visible light image and the infrared image to obtain a fused image.

[0095] The processor can access the information and application programs stored in the memory via the transmission device to perform the following steps: calculate the credibility of image features, the credibility of speech features, and the credibility of operation behavior features respectively; determine the weights of image features, speech features, and operation behavior features based on the credibility of image features, speech features, and operation behavior features; and perform weighted fusion processing based on image features, image feature weights, speech features, speech feature weights, operation behavior features, and operation behavior feature weights to obtain fused data.

[0096] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: compare the fused data with a preset threshold; if the fused data is greater than the preset threshold, then the target transaction is considered to have an anomaly as the judgment result; if the fused data is less than or equal to the preset threshold, then the target transaction is considered to have no anomaly as the judgment result.

[0097] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: after determining that the target transaction is abnormal, calculate the difference between the fused data and the preset threshold, and compare the difference with the warning threshold; if the difference is greater than the warning threshold, execute the first warning strategy; if the difference is less than or equal to the warning threshold, execute the second warning strategy, wherein the warning processing flow corresponding to the first warning strategy and the second warning strategy is different.

[0098] The processor can access information and applications stored in the memory via a transmission device to execute the following steps: Before acquiring multimodal data corresponding to the target transaction, images are captured using the ATM's camera and infrared camera to obtain visible light and infrared images, and image data is generated based on the visible light and infrared images; voice data is acquired using the ATM's voice acquisition module; light intensity is acquired using the ATM's ambient light sensor, and operation behavior data is acquired using the ATM's operation behavior acquisition module, and environmental data is generated based on the light intensity and operation behavior data.

[0099] Those skilled in the art will understand that Figure 5 The structure shown is for illustrative purposes only. Electronic devices can also be smartphones, tablets, handheld computers, mobile internet devices (MIDs), PADs, and other terminal devices. Figure 5 This does not limit the structure of the aforementioned electronic device. For example, electronic devices may also include components that are more... Figure 5 The more or fewer components shown (such as network interfaces, display devices, etc.), or having the same Figure 5 The different configurations shown.

[0100] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0101] Example 4

[0102] Embodiments of this application also provide a computer-readable storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the abnormal transaction identification method provided in Embodiment 1.

[0103] Optionally, in this embodiment, the storage medium may be located in any computer terminal in a group of computer terminals in a computer network, or in any mobile terminal in a group of mobile terminals.

[0104] This application also provides a computer program product, which, when executed on a data processing device, is suitable for performing steps of a method for identifying abnormal transactions.

[0105] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0106] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0107] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0108] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0109] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0110] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0111] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for identifying abnormal transactions, characterized in that, Applications in ATMs include: Acquire multimodal data corresponding to the target transaction, and determine the target transaction scenario corresponding to the target transaction based on the multimodal data, wherein the multimodal data includes image data, voice data, and environmental data; When the target transaction scenario is the first transaction scenario, feature extraction processing is performed on the image data, the voice data, and the operation behavior data in the environmental data to obtain image features, voice features, and operation behavior features. The first transaction scenario represents a scenario where the light intensity is greater than the light threshold and the facial occlusion area of ​​the transaction object initiating the target transaction is greater than the occlusion threshold. The image features, voice features, and operation behavior features are weighted and fused to obtain fused data, wherein the fused data is used to characterize the degree of abnormality of the target transaction; Based on the fused data and the preset threshold corresponding to the target transaction scenario, it is determined whether the target transaction is abnormal, and a judgment result is obtained.

2. The method according to claim 1, characterized in that, When the target transaction scenario is the first transaction scenario, feature extraction processing is performed on the operational behavior data in the image data, the voice data, and the environmental data respectively to obtain image features, voice features, and operational behavior features, including: A fused image is generated based on the visible light image and infrared image in the image data, and feature extraction processing is performed on the fused image to obtain the image features; The speech data is subjected to noise reduction processing to obtain noise-reduced speech data, and the noise-reduced speech data is subjected to feature extraction processing to obtain the speech features; The operational behavior data is statistically analyzed to obtain statistical data, and the statistical data is then processed by feature extraction to obtain the operational behavior features.

3. The method according to claim 2, characterized in that, Generating a fused image based on the visible light image and infrared image in the image data includes: The visible light image is subjected to light correction processing to obtain a corrected visible light image; The corrected visible light image is enhanced using a preset model to obtain an enhanced visible light image. The enhanced visible light image and the infrared image are weighted and fused to obtain the fused image.

4. The method according to claim 1, characterized in that, The fused data is obtained by weighted fusion processing based on the image features, the speech features, and the operation behavior features, resulting in the following: The credibility of the image features, the credibility of the speech features, and the credibility of the operation behavior features are calculated respectively. Based on the reliability of the image features, the reliability of the speech features, and the reliability of the operation behavior features, the weights of the image features, the speech features, and the operation behavior features are determined. The fused data is obtained by performing weighted fusion processing based on the image features, the weights of the image features, the speech features, the weights of the speech features, the operation behavior features, and the weights of the operation behavior features.

5. The method according to claim 1, characterized in that, Based on the fused data and the preset threshold corresponding to the target transaction scenario, it is determined whether the target transaction is abnormal, and the determination result includes: The fused data is compared with the preset threshold; If the fused data is greater than the preset threshold, then the judgment result is that the target transaction is abnormal; If the fused data is less than or equal to the preset threshold, then the judgment result is that the target transaction does not have any anomalies.

6. The method according to claim 5, characterized in that, After determining that the target transaction is abnormal, the method further includes: Calculate the difference between the fused data and the preset threshold, and compare the difference with the warning threshold; If the difference is greater than the warning threshold, then the first warning strategy is executed; If the difference is less than or equal to the warning threshold, a second warning strategy is executed, wherein the warning processing procedures corresponding to the first warning strategy and the second warning strategy are different.

7. The method according to claim 1, characterized in that, Before acquiring the multimodal data corresponding to the target transaction, the method further includes: Images are acquired using the camera and infrared camera of the automated teller machine to obtain visible light images and infrared images, and image data is generated based on the visible light images and infrared images; The voice data is obtained by voice acquisition module of the ATM. The ambient light sensor of the ATM collects light intensity, and the operation behavior collection module of the ATM collects operation behavior data. The environmental data is then generated based on the light intensity and the operation behavior data.

8. A device for identifying abnormal transactions, characterized in that, Applications in ATMs include: An acquisition unit is used to acquire multimodal data corresponding to a target transaction and determine the target transaction scenario corresponding to the target transaction based on the multimodal data, wherein the multimodal data includes image data, voice data, and environmental data; The first processing unit is configured to perform feature extraction processing on the image data, the voice data, and the operation behavior data in the environment data respectively when the target transaction scenario is the first transaction scenario, to obtain image features, voice features, and operation behavior features, wherein the first transaction scenario represents a scenario where the light intensity is greater than the light threshold and the face occlusion area of ​​the transaction object initiating the target transaction is greater than the occlusion threshold. The second processing unit is used to perform weighted fusion processing on the image features, the voice features and the operation behavior features to obtain fused data, wherein the fused data is used to characterize the degree of abnormality of the target transaction; The third processing unit is used to determine whether there is an anomaly in the target transaction based on the fused data and the preset threshold corresponding to the target transaction scenario, and to obtain a judgment result.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored executable program, wherein, when the executable program is executed, it controls the device on which the computer-readable storage medium is located to perform the abnormal transaction identification method according to any one of claims 1 to 7.

10. An electronic device, characterized in that, include: Memory, which stores executable programs; A processor for running the program, wherein the program, when running, executes the method for identifying abnormal transactions as described in any one of claims 1 to 7.

Images