Signature data processing method and device, computer equipment and storage medium

Abstract

The invention belongs to the technical field of artificial intelligence, and relates to a signature data processing method and device, computer equipment and a storage medium, and the method comprises the steps: obtaining image data collected in a signature process; detecting the image data based on a terminal detection model to obtain a terminal detection result; based on the terminal detection result, calling an action detection model to perform signature action detection on the image data to obtain a signature action detection result; if the signature action detection result is matched with the signature action mode, calling a track detection model to perform signature track detection on the image data based on the terminal detection result to obtain a signature track detection result; if the signature track detection result is matched with the signature track mode, carrying out signature extraction on the image data to obtain signature data; and comparing the signature data based on the signature recognition model and the signature information to obtain a signature comparison result. In addition, the signature comparison result can be stored in the block chain. According to the method and the device, the comprehensiveness and the comparison accuracy of authenticity verification of the signature are improved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence development technology and financial technology, and in particular to methods, devices, computer equipment and storage media for processing signature data. Background Art

[0002] In the financial field, there are scenarios where the authenticity of the user's signature needs to be checked, such as in the dual recording process of financial and insurance products (i.e., "audio and video dual recording"), where the user's signature is crucial. It not only represents the user's confirmation of the contract terms, but is also an important means to ensure the authenticity and legality of the transaction. However, in existing technologies, the authenticity verification of user signatures mostly uses simple signature action detection or signature comparison methods. These processing methods are relatively simple, resulting in low accuracy and difficulty in fully meeting the high security requirements of the financial and insurance industries. Summary of the invention

[0003] The purpose of the embodiments of the present application is to propose a method, apparatus, computer device and storage medium for processing signature data to solve the technical problem that the existing authenticity verification of user signatures mostly adopts a relatively single processing method, resulting in low accuracy.

[0004] In order to solve the above technical problems, the present application embodiment provides a method for processing signature data, which adopts the following technical solution:

[0005] Obtaining image data corresponding to the user collected during the signing process;

[0006] Performing detection processing on the image data based on a preset terminal detection model to obtain a corresponding terminal detection result;

[0007] Based on the terminal detection result, calling a preset action detection model to perform signature action detection on the image data to obtain a corresponding signature action detection result;

[0008] If the signature action detection result matches the preset signature action pattern, based on the terminal detection result, calling a preset trajectory detection model to perform signature trajectory detection on the image data to obtain a corresponding signature trajectory detection result;

[0009] If the signature trajectory detection result matches the preset signature trajectory pattern, the signature is extracted from the image data to obtain corresponding signature data;

[0010] The signature data is compared with the pre-stored signature information based on a preset signature recognition model to obtain a signature comparison result corresponding to the signature data.

[0011] Furthermore, the step of calling a preset action detection model to perform signature action detection on the image data based on the terminal detection result to obtain a corresponding signature action detection result specifically includes:

[0012] Determine a first position of the hand in the image data based on the terminal detection result, and acquire a hand area image corresponding to the first position;

[0013] Preprocessing the hand region image to obtain a corresponding first image;

[0014] Performing hand key point detection on the first image based on a preset hand key point detection algorithm to obtain corresponding target key points;

[0015] Extracting features of the target key points to obtain corresponding target features;

[0016] Classify the target feature based on the action detection model to obtain a corresponding first classification result;

[0017] The signature action detection result is generated based on the first classification result.

[0018] Furthermore, the step of calling a preset trajectory detection model to perform signature trajectory detection on the image data based on the terminal detection result to obtain a corresponding signature trajectory detection result specifically includes:

[0019] Determine a second position of the signature area in the image data based on the terminal detection result, and obtain a signature image corresponding to the second position;

[0020] Preprocessing the signature image to obtain a corresponding second image;

[0021] Performing classification processing on the second image based on the trajectory detection model to obtain a corresponding second classification result;

[0022] The signature trajectory detection result is generated based on the second classification result.

[0023] Furthermore, the step of extracting the signature from the image data to obtain the corresponding signature data specifically includes:

[0024] Obtaining a signature position in the image data;

[0025] Cropping the image data based on the signature position to obtain a corresponding third image;

[0026] Performing edge detection processing on the third image based on a preset edge detection algorithm to obtain a corresponding fourth image;

[0027] The signature data is generated based on the fourth image.

[0028] Furthermore, the step of comparing the signature data based on a preset signature recognition model with pre-stored signature information to obtain a signature comparison result corresponding to the signature data specifically includes:

[0029] Obtaining a specified form corresponding to the signature information;

[0030] Performing recognition processing on the signature data based on the signature recognition model to obtain designated information corresponding to the designated form;

[0031] Obtaining a specified comparison strategy corresponding to the specified form;

[0032] The designated information and the signature information are compared based on the designated comparison strategy to obtain a signature comparison result corresponding to the signature data.

[0033] Furthermore, the step of comparing the designated information with the signature information based on the designated comparison strategy to obtain a signature comparison result corresponding to the signature data specifically includes:

[0034] If the specified form of the signature information is a feature vector form, calling a preset similarity algorithm;

[0035] Calculating the similarity between the specified information and the signature information based on the similarity algorithm;

[0036] Determining whether the similarity is greater than a preset similarity threshold;

[0037] If so, the generated signature is the first comparison result of the real signature;

[0038] If not, a second comparison result is generated indicating that the signature is a forged signature.

[0039] Furthermore, before the step of performing detection processing on the image data based on a preset terminal detection model to obtain a corresponding terminal detection result, the method further includes:

[0040] Call the preset initial detection model and obtain the preset engineering transformation strategy;

[0041] Based on the engineering transformation strategy, the initial detection model is transformed to obtain a corresponding first detection model;

[0042] Obtaining pre-collected initial sample data related to terminal detection;

[0043] Performing data enhancement processing on the initial sample data to obtain corresponding target sample data;

[0044] Training the first detection model based on the target sample data to obtain a second detection model that meets the preset construction requirements;

[0045] The second detection model is used as the terminal detection model.

[0046] In order to solve the above technical problems, the embodiment of the present application also provides a signature data processing device, which adopts the following technical solution:

[0047] A first acquisition module, used to acquire image data corresponding to the user collected during the signing process;

[0048] A first detection module, used to perform detection processing on the image data based on a preset terminal detection model to obtain a corresponding terminal detection result;

[0049] A second detection module is used to call a preset action detection model to perform signature action detection on the image data based on the terminal detection result to obtain a corresponding signature action detection result;

[0050] A third detection module, configured to, if the signature action detection result matches a preset signature action pattern, call a preset trajectory detection model to perform signature trajectory detection on the image data based on the terminal detection result to obtain a corresponding signature trajectory detection result;

[0051] An extraction module, configured to extract the signature from the image data to obtain corresponding signature data if the signature trajectory detection result matches a preset signature trajectory pattern;

[0052] The comparison module is used to compare the signature data based on a preset signature recognition model and pre-stored signature information to obtain a signature comparison result corresponding to the signature data.

[0053] In order to solve the above technical problems, the embodiment of the present application further provides a computer device, which adopts the following technical solution:

[0054] Obtaining image data corresponding to the user collected during the signing process;

[0055] Performing detection processing on the image data based on a preset terminal detection model to obtain a corresponding terminal detection result;

[0056] Based on the terminal detection result, calling a preset action detection model to perform signature action detection on the image data to obtain a corresponding signature action detection result;

[0057] If the signature action detection result matches the preset signature action pattern, based on the terminal detection result, calling a preset trajectory detection model to perform signature trajectory detection on the image data to obtain a corresponding signature trajectory detection result;

[0058] If the signature trajectory detection result matches the preset signature trajectory pattern, the signature is extracted from the image data to obtain corresponding signature data;

[0059] The signature data is compared with the pre-stored signature information based on a preset signature recognition model to obtain a signature comparison result corresponding to the signature data.

[0060] In order to solve the above technical problems, the embodiment of the present application further provides a computer-readable storage medium, which adopts the following technical solution:

[0061] Obtaining image data corresponding to the user collected during the signing process;

[0062] Performing detection processing on the image data based on a preset terminal detection model to obtain a corresponding terminal detection result;

[0063] Based on the terminal detection result, calling a preset action detection model to perform signature action detection on the image data to obtain a corresponding signature action detection result;

[0064] If the signature action detection result matches the preset signature action pattern, based on the terminal detection result, calling a preset trajectory detection model to perform signature trajectory detection on the image data to obtain a corresponding signature trajectory detection result;

[0065] If the signature trajectory detection result matches the preset signature trajectory pattern, the signature is extracted from the image data to obtain corresponding signature data;

[0066] The signature data is compared with the pre-stored signature information based on a preset signature recognition model to obtain a signature comparison result corresponding to the signature data.

[0067] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0068] The present application first obtains image data corresponding to the user collected during the signing process; then detects and processes the image data based on a preset terminal detection model to obtain a corresponding terminal detection result; then, based on the terminal detection result, calls a preset action detection model to perform signature action detection on the image data to obtain a corresponding signature action detection result; if the signature action detection result matches the preset signature action pattern, then, based on the terminal detection result, calls a preset trajectory detection model to perform signature trajectory detection on the image data to obtain a corresponding signature trajectory detection result; if the signature trajectory detection result matches the preset signature trajectory pattern, then performs signature extraction on the image data to obtain the corresponding signature data; finally, the signature data is compared and processed based on the preset signature recognition model and the pre-stored signature information to obtain a signature comparison result corresponding to the signature data. After obtaining the image data corresponding to the user collected during the signing process, the present application performs terminal detection, signature action detection, signature trajectory detection and signature comparison on the image data by combining the use of terminal detection model, action detection model, trajectory detection model and signature recognition model to generate the final signature comparison result, thereby effectively improving the comprehensiveness of the signature authenticity verification detection and improving the accuracy of the generated signature comparison result. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] In order to more clearly illustrate the scheme in the present application, a brief introduction is given below to the drawings required for use in the description of the embodiments of the present application. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0070] Figure 1 is an exemplary system architecture diagram to which the present application may be applied;

[0071] Figure 2 A flowchart of an embodiment of a method for processing signature data according to the present application;

[0072] Figure 3 is a schematic structural diagram of an embodiment of a signature data processing device according to the present application;

[0073] Figure 4 It is a structural diagram of an embodiment of a computer device according to the present application. DETAILED DESCRIPTION

[0074] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of the present application; the terms used in the specification of the application herein are only for the purpose of describing specific embodiments and are not intended to limit the present application; the terms "including" and "having" and any variations thereof in the specification and claims of the present application and the above-mentioned drawings are intended to cover non-exclusive inclusions. The terms "first", "second", etc. in the specification and claims of the present application or the above-mentioned drawings are used to distinguish different objects, not to describe a specific order.

[0075] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0076] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings.

[0077] like Figure 1 As shown, the system architecture 100 may include a terminal device 101, a network 102 and a server 103. The terminal device 101 may be a laptop 1011, a tablet computer 1012 or a mobile phone 1013. The network 102 is used to provide a medium for a communication link between the terminal device 101 and the server 103. The network 102 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0078] The user can use the terminal device 101 to interact with the server 103 through the network 102 to receive or send messages, etc. Various communication client applications can be installed on the terminal device 101, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.

[0079] The terminal device 101 can be any electronic device with a display screen and supporting web browsing. In addition to the laptop computer 1011, tablet computer 1012 or mobile phone 1013, the terminal device 101 can also be an e-book reader, an MP3 player (Moving Picture Experts Group Audio Layer III), an MP4 (Moving Picture Experts Group Audio Layer IV), a laptop computer, a desktop computer, etc.

[0080] The server 103 may be a server that provides various services, such as a background server that provides support for a web page displayed on the terminal device 101 .

[0081] It should be noted that the signature data processing method provided in the embodiment of the present application is generally executed by a server / terminal device, and accordingly, the signature data processing device is generally set in the server / terminal device.

[0082] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0083] Continue to refer Figure 2 , shows a flow chart of an embodiment of a method for processing signature data according to the present application. According to different requirements, the order of the steps in the flow chart can be changed, and some steps can be omitted. The method for processing signature data provided in the embodiment of the present application can be applied to any scenario where signature comparison is required, and the method for processing signature data can be applied to products in these scenarios, for example, signature comparison in the field of finance and insurance. The method for processing signature data comprises the following steps:

[0084] Step S201, obtaining image data corresponding to the user collected during the signing process.

[0085] In this embodiment, the signature data processing method is executed on the electronic device (eg Figure 1The server / terminal device shown in the figure) can obtain image data through a wired connection or a wireless connection. It should be noted that the above-mentioned wireless connection methods may include but are not limited to 3G / 4G / 5G connection, WiFi connection, Bluetooth connection, WiMAX connection, Zigbee connection, UWB (ultra wideband) connection, and other wireless connection methods currently known or to be developed in the future. The executor of the present application is a system with a signature comparison function. The present application can be applied to scenarios in the financial field where the authenticity of the user's signature needs to be detected, for example, it can be applied to business scenarios where the user signs during the dual recording (audio and video dual recording) process of financial and insurance products. Correspondingly, the above-mentioned image data may be image data collected during the user's signature process and containing data signed by the user on the terminal (mobile phone).

[0086] Step S202: Detect and process the image data based on a preset terminal detection model to obtain a corresponding terminal detection result.

[0087] In this embodiment, the above image data can be input into the terminal detection model, so that the terminal detection model can detect and process the image data to output the corresponding terminal detection result. Among them, the terminal can specifically refer to a mobile phone, and the terminal detection model is also the corresponding mobile phone detection model. The terminal detection model can accurately identify the mobile phone area in the image data, including key information such as the position and size of the mobile phone. Mobile phone detection can automatically identify and process the mobile phone part in the image data, reduce the steps of manually locating the signature area, and improve the automation of the verification process. Specifically, the detection frame returned by the mobile phone detection can be positioned to extract the image of the signature area from the image containing the mobile phone, so that by accurately locating the mobile phone area, the image part containing the signature can be further cropped, thereby reducing background interference and improving the accuracy of signature trajectory detection. In addition, the specific construction process of the above terminal detection model will be further described in detail in the subsequent specific embodiments of this application, and will not be elaborated on here.

[0088] Step S203: Based on the terminal detection result, a preset action detection model is called to perform signature action detection on the image data to obtain a corresponding signature action detection result.

[0089] In this embodiment, the above-mentioned specific implementation process of calling the preset action detection model to perform signature action detection on the image data based on the terminal detection result to obtain the corresponding signature action detection result will be further described in detail in the subsequent specific embodiments of this application, and will not be elaborated on here. Among them, if the signature action detection result does not match the preset signature action mode, a signature verification result indicating that the signature is a forged signature is directly generated.

[0090] Step S204: If the signature action detection result matches the preset signature action pattern, based on the terminal detection result, a preset trajectory detection model is called to perform signature trajectory detection on the image data to obtain a corresponding signature trajectory detection result.

[0091] In this embodiment, the specific implementation process of calling the preset trajectory detection model to perform signature trajectory detection on the image data based on the terminal detection result to obtain the corresponding signature trajectory detection result will be further described in detail in the subsequent specific embodiments of this application, and will not be elaborated on here. Among them, if the signature trajectory detection result does not match the preset signature trajectory pattern, a signature verification result indicating that the signature is a forged signature is directly generated.

[0092] Step S205: If the signature trajectory detection result matches the preset signature trajectory pattern, the signature is extracted from the image data to obtain corresponding signature data.

[0093] In this embodiment, the specific implementation process of extracting the signature from the image data to obtain the corresponding signature data will be described in further detail in subsequent specific embodiments of the present application and will not be elaborated on here.

[0094] Step S206: performing a comparison process on the signature data based on a preset signature recognition model and pre-stored signature information to obtain a signature comparison result corresponding to the signature data.

[0095] In this embodiment, the above-mentioned specific implementation process of comparing the signature data based on the preset signature recognition model and the pre-stored signature information to obtain the signature comparison result corresponding to the signature data will be further described in detail in the subsequent specific embodiments of this application and will not be elaborated in detail here.

[0096] The present application first obtains image data corresponding to the user collected during the signing process; then detects and processes the image data based on a preset terminal detection model to obtain a corresponding terminal detection result; then, based on the terminal detection result, calls a preset action detection model to perform signature action detection on the image data to obtain a corresponding signature action detection result; if the signature action detection result matches the preset signature action pattern, then, based on the terminal detection result, calls a preset trajectory detection model to perform signature trajectory detection on the image data to obtain a corresponding signature trajectory detection result; if the signature trajectory detection result matches the preset signature trajectory pattern, then performs signature extraction on the image data to obtain the corresponding signature data; finally, the signature data is compared and processed based on the preset signature recognition model and the pre-stored signature information to obtain a signature comparison result corresponding to the signature data. After obtaining the image data corresponding to the user collected during the signing process, the present application performs terminal detection, signature action detection, signature trajectory detection and signature comparison on the image data by combining the use of terminal detection models, action detection models, trajectory detection models and signature recognition models to generate the final signature comparison result, thereby effectively improving the comprehensiveness of the signature authenticity verification detection and improving the accuracy of the generated signature comparison result.

[0097] In some optional implementations, step S203 includes the following steps:

[0098] Based on the terminal detection result, a first position of the hand in the image data is determined, and a hand area image corresponding to the first position is acquired.

[0099] In this embodiment, the terminal detection result can be obtained by analyzing the output of the terminal detection model, and the first position of the hand in the image data can be determined based on the terminal detection result, and then the matching hand area image can be extracted from the above image data based on the first position.

[0100] The hand region image is preprocessed to obtain a corresponding first image.

[0101] In this embodiment, the preprocessing may include cropping and scaling the hand region image to facilitate key point detection, wherein the hand region is ensured to be complete and clear during cropping, and the proportion of the hand region is kept unchanged during scaling.

[0102] Hand key point detection is performed on the first image based on a preset hand key point detection algorithm to obtain corresponding target key points.

[0103] In this embodiment, the hand key point detection model corresponding to the above-mentioned hand key point detection algorithm can be loaded by using the Opencv DNN framework. The above-mentioned hand key point detection model has been trained and has a high detection accuracy. Specifically, it can be achieved by inputting the first image into the hand key point detection model and parsing the output of the hand key point detection model. The detected target key points include key parts such as finger joints.

[0104] Feature extraction is performed on the target key points to obtain corresponding target features.

[0105] In this embodiment, the movement trajectory and speed of the finger can be calculated based on the detected target key points to obtain the above target features. This can be achieved by tracking the position changes of the target key points. For example, parameters such as the displacement and speed of the key points between adjacent frames can be calculated. The extracted target features are representative and easy to classify, which helps to improve the accuracy and robustness of the action classification process.

[0106] The target feature is classified based on the action detection model to obtain a corresponding first classification result.

[0107] In this embodiment, the above-mentioned action detection model can specifically adopt a MobileNetV3 classification model that has been trained and has a high classification accuracy. The extracted target features can be input into the MobileNetV3 classification model for classification processing, and the output of the model can be analyzed to determine whether the movement of the finger constitutes a signature action. Among them, the classification result is a binary classification (signature / non-signature). In addition, according to the classification result, the authenticity and consistency of the signature can be further judged. For example, if the signature action is detected to match the predefined signature action pattern, the signature can be considered to be authentic; otherwise, the signature can be considered to be forged or inconsistent.

[0108] The signature action detection result is generated based on the first classification result.

[0109] In this embodiment, the first classification result can be directly generated as the above-mentioned signature action detection result.

[0110] This application determines the first position of the hand in the image data based on the terminal detection result, and obtains the hand area image corresponding to the first position; then pre-processes the hand area image to obtain the corresponding first image; then performs hand key point detection on the first image based on a preset hand key point detection algorithm to obtain the corresponding target key point; subsequently performs feature extraction on the target key point to obtain the corresponding target feature; further classifies the target feature based on the action detection model to obtain the corresponding first classification result; and finally generates the signature action detection result based on the first classification result. Based on the use of terminal detection results, hand key point detection algorithm and action detection model, this application can quickly and accurately complete the signature action detection processing of image data, improve the processing efficiency and processing accuracy of signature action detection, and effectively ensure the accuracy of the generated signature action detection result.

[0111] In some optional implementations of this embodiment, step S204 includes the following steps:

[0112] A second position of the signature area in the image data is determined based on the terminal detection result, and a signature image corresponding to the second position is acquired.

[0113] In this embodiment, the precise position of the signature area in the original image data, i.e., the second position, can be determined based on the terminal detection result. The second position usually includes a rectangular frame, which indicates the approximate range of the signature on the image. Further, the located second position can be used to crop the signature area from the image data, and ensure that the cropped image (signature image) only contains the signature part and maintains the original proportion and clarity of the signature.

[0114] The signature image is preprocessed to obtain a corresponding second image.

[0115] In this embodiment, the obtained signature image can be grayed to reduce the interference of color information, and further subjected to noise reduction processing to eliminate noise and miscellaneous points in the image, thereby improving the accuracy of subsequent processing.

[0116] The second image is classified based on the trajectory detection model to obtain a corresponding second classification result.

[0117] In this embodiment, the trajectory detection model is a MobileNetV3 model (or a similar model) that has been trained with trajectory detection function. The model can identify trajectory features in the preprocessed image and output the prediction result of the trajectory type (such as a straight line, a curve, etc.).

[0118] The signature trajectory detection result is generated based on the second classification result.

[0119] In this embodiment, the second classification result can be directly used as the signature trajectory detection result. The authenticity and consistency of the signature can be further determined based on the signature trajectory detection result. For example, if the signature trajectory detection result is detected to match the predefined signature trajectory pattern, the signature can be considered to be authentic; otherwise, the signature can be considered to be forged or inconsistent.

[0120] In addition, signature trajectory data in real scenarios can be collected and transformed by blurring, rotating, etc. to increase data diversity and model generalization. Ensure that the enhanced data can cover all possible signature trajectories and scenarios. The enhanced data is then used to train and optimize the trajectory detection model. By adjusting the parameters and structure of the model, its accuracy and robustness in various scenarios can be improved. Specifically, this may include adjusting hyperparameters such as learning rate, optimizer, loss function, and trying different network structures and numbers of layers.

[0121] The present application determines the second position of the signature area in the image data based on the terminal detection result, and obtains the signature image corresponding to the second position; then pre-processes the signature image to obtain the corresponding second image; then classifies the second image based on the trajectory detection model to obtain the corresponding second classification result; and subsequently generates the signature trajectory detection result based on the second classification result. The present application determines the second position of the signature area in the image data based on the terminal detection result, and obtains the signature image corresponding to the second position; then pre-processes the signature image to obtain the second image, and then based on the use of the trajectory detection model, it can quickly and accurately complete the signature trajectory detection processing for the image data, improve the processing efficiency and accuracy of the signature trajectory detection, and effectively ensure the accuracy of the generated signature trajectory detection result.

[0122] In some optional implementations, step S205 includes the following steps:

[0123] The signature position in the image data is obtained.

[0124] In this embodiment, the above-mentioned acquisition of the signature position in the image data can refer to the above-mentioned processing process of determining the second position of the signature area in the image data based on the terminal detection result, and no further details are given here.

[0125] The image data is cropped based on the signature position to obtain a corresponding third image.

[0126] In this embodiment, the original image data can be cropped according to the obtained signature position to obtain the smallest rectangular frame containing the signature, i.e. the third image mentioned above. In the cropping process, it is ensured that the cropped image only contains the signature part and the signature is clearly identifiable.

[0127] Perform edge detection processing on the third image based on a preset edge detection algorithm to obtain a corresponding fourth image.

[0128] In this embodiment, edge detection can be performed on the cropped third image by using an edge detection algorithm of OPENCV (such as Canny edge detection), so that a clear outline of the signature can be extracted to obtain a corresponding fourth image.

[0129] The signature data is generated based on the fourth image.

[0130] In this embodiment, the fourth image can be directly used as the signature data.

[0131] The present application obtains the signature position in the image data; then crops the image data based on the signature position to obtain a corresponding third image; then performs edge detection processing on the third image based on a preset edge detection algorithm to obtain a corresponding fourth image; and subsequently generates the signature data based on the fourth image. The present application crops the image data according to the signature position in the acquired image data to obtain a corresponding third image, and then performs edge detection processing on the third image based on the use of an edge detection algorithm, thereby achieving fast and accurate completion of the signature extraction processing for the image data, improving the processing efficiency and accuracy of the signature extraction, and effectively ensuring the accuracy of the generated signature data.

[0132] In some optional implementations, step S206 includes the following steps:

[0133] Acquire a specified format corresponding to the signature information.

[0134] In this embodiment, the signature information is pre-stored archive information used for the user's signature verification process. The signature information may be in a text form or a feature vector form.

[0135] The signature data is identified based on the signature identification model to obtain designated information corresponding to the designated form.

[0136] In this embodiment, the signature recognition model may be a recognition model based on ResNet50. By using the signature recognition model to perform recognition processing on the signature data, the signature data may be converted into a feature vector form or a text form.

[0137] Get a specified comparison strategy corresponding to the specified format.

[0138] In this embodiment, the comparison strategy corresponding to the text form includes: directly performing text comparison. The comparison strategy corresponding to the feature vector form includes: using a similarity algorithm to calculate the similarity between the recognition result of the signature recognition model and the archived information. Subsequently, the authenticity and consistency of the signature are judged according to the similarity threshold. If the similarity is higher than the threshold, the signature is considered to be authentic; otherwise, the signature is considered to be forged.

[0139] The designated information and the signature information are compared based on the designated comparison strategy to obtain a signature comparison result corresponding to the signature data.

[0140] In this embodiment, the specific implementation process of comparing the specified information with the signature information based on the specified comparison strategy to obtain the signature comparison result corresponding to the signature data will be further described in detail in subsequent specific embodiments of the present application and will not be elaborated on here.

[0141] The present application obtains the designated form corresponding to the signature information; then identifies and processes the signature data based on the signature recognition model to obtain the designated information corresponding to the designated form; then obtains the designated comparison strategy corresponding to the designated form; and subsequently compares and processes the designated information with the signature information based on the designated comparison strategy to obtain a signature comparison result corresponding to the signature data. The present application identifies and processes the signature data based on the use of the signature recognition model to obtain the designated information corresponding to the designated form of the signature information, and then compares and processes the designated information with the signature information based on the use of the designated comparison strategy corresponding to the designated form, thereby improving the processing efficiency and processing intelligence of the comparison processing and ensuring the data accuracy of the obtained signature comparison result.

[0142] In some optional implementations of this embodiment, the comparing the specified information with the signature information based on the specified comparison strategy to obtain a signature comparison result corresponding to the signature data includes the following steps:

[0143] If the specified form of the signature information is a feature vector form, a preset similarity algorithm is called.

[0144] In this embodiment, there is no specific limitation on the selection of the above similarity algorithm, for example, cosine similarity may be used.

[0145] The similarity between the designated information and the signature information is calculated based on the similarity algorithm.

[0146] In this embodiment, the similarity between the designated information and the signature information may be calculated by using the similarity algorithm to obtain the corresponding similarity.

[0147] It is determined whether the similarity is greater than a preset similarity threshold.

[0148] In this embodiment, there is no specific limitation on the value of the similarity threshold, which can be set according to actual comparison requirements, for example, it can be set to 0.88.

[0149] If so, the generated signature is the first comparison result of the authentic signature.

[0150] In this embodiment, if the similarity is higher than the similarity threshold, the signature is considered to be authentic, and a first comparison result indicating that the signature is an authentic signature is generated.

[0151] If not, a second comparison result is generated indicating that the signature is a forged signature.

[0152] In this embodiment, if the similarity is lower than the similarity threshold, the signature is considered inconsistent or forged, and a second comparison result indicating that the signature is a forged signature is generated.

[0153] If the present application detects that the specified form of the signature information is in the form of a feature vector, the preset similarity algorithm is called; then the similarity between the specified information and the signature information is calculated based on the similarity algorithm; subsequently, it is determined whether the similarity is greater than a preset similarity threshold; if so, a first comparison result is generated indicating that the signature is a genuine signature; if not, a second comparison result is generated indicating that the signature is a forged signature. When the present application detects that the specified form of the signature information is in the form of a feature vector, the similarity algorithm is automatically and intelligently called to calculate the similarity between the specified information and the signature information, and then the obtained similarity is compared with the preset similarity threshold, and then the signature comparison result corresponding to the signature data can be accurately generated based on the obtained comparison result, thereby ensuring the accuracy of the generated signature comparison result.

[0154] In some optional implementations of this embodiment, before step S202, the electronic device may further perform the following steps:

[0155] Call the preset initial detection model and obtain the preset engineering transformation strategy.

[0156] In this embodiment, the initial detection model may specifically adopt a self-developed model based on the YOLO framework, which has sufficient generalization capability and can adapt to the detection requirements of different terminal models, angles and positions. The terminal may specifically refer to a mobile phone.

[0157] The initial detection model is transformed based on the engineering transformation strategy to obtain a corresponding first detection model.

[0158] In this embodiment, the above-mentioned engineering transformation strategy may include a preprocessing optimization strategy, a post-processing optimization strategy, and a model quantization strategy. Among them, the preprocessing optimization strategy includes: preprocessing the input image according to the actual application scenario, such as adjusting the image size, normalization, etc., to improve the detection effect of the model. The post-processing optimization strategy includes: using the WBF (Weighted Boxes Fusion) algorithm instead of the NMS (Non-Maximum Suppression) algorithm to reduce missed detection and false detection of overlapping targets. The WBF algorithm fuses multiple detection frames by considering the weight and overlap of the detection frame, thereby improving the detection accuracy. The model quantization strategy includes: applying model quantization technology to convert the model from floating point to INT8 integer. This can significantly reduce the computational complexity and memory usage of the model while maintaining a high detection accuracy. The quantized model is more suitable for running on resource-constrained devices.

[0159] Obtain pre-collected initial sample data related to terminal detection.

[0160] In this embodiment, image data of various mobile phone models, angles and positions can be collected from actual signature scenes. Ensure that the data covers a variety of backgrounds, lighting conditions and signature postures to improve the generalization ability of the model. Then, the collected data is annotated, including information such as the location and model of the mobile phone, so as to obtain the corresponding initial sample data.

[0161] Perform data enhancement processing on the initial sample data to obtain corresponding target sample data.

[0162] In this embodiment, the collected initial sample data can be rotated, scaled, cropped, flipped, etc. to increase the diversity of the data. This helps the model better adapt to the needs of mobile phone detection in different scenarios. In addition, special data enhancement processing is performed for the difficulties in mobile phone detection, such as horizontal screen and partial mobile phone. For example, the detection of mobile phones at different angles and positions can be simulated by cropping and scaling.

[0163] The first detection model is trained based on the target sample data to obtain a second detection model that meets the preset construction requirements.

[0164] In this embodiment, the first detection model can be trained using the target sample data to construct a second detection model having a terminal detection function and meeting the construction requirements, wherein the construction requirements can be determined according to actual model construction requirements.

[0165] The second detection model is used as the terminal detection model.

[0166] The present application calls a preset initial detection model and obtains a preset engineering transformation strategy; then transforms the initial detection model based on the engineering transformation strategy to obtain a corresponding first detection model; then obtains pre-collected initial sample data related to terminal detection; and performs data enhancement processing on the initial sample data to obtain corresponding target sample data; subsequently, the first detection model is trained based on the target sample data to obtain a second detection model that meets the preset construction requirements, and the second detection model is used as the terminal detection model, thereby achieving rapid and accurate completion of the terminal detection model construction process, improving the terminal detection model construction efficiency, and ensuring the model effect of the obtained terminal detection model.

[0167] In some optional implementations, the user information obtained is subject to the user's consent and complies with relevant laws and policies.

[0168] In addition, the non-Company software tools or components that appear in the embodiments of the present application are merely examples and do not represent actual use.

[0169] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0170] It should be emphasized that in order to further ensure the privacy and security of the above-mentioned signature comparison results, the above-mentioned signature comparison results can also be stored in a node of a blockchain.

[0171] The blockchain referred to in this application is a new application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, encryption algorithm, etc. Blockchain is essentially a decentralized database, a string of data blocks generated by cryptographic methods. Each data block contains a batch of network transaction information, which is used to verify the validity of its information (anti-counterfeiting) and generate the next block. Blockchain can include the underlying blockchain platform, platform product service layer, and application service layer.

[0172] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0173] AI basic technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. AI software technologies mainly include computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0174] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through computer-readable instructions, and the computer-readable instructions 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-mentioned methods. Among them, the aforementioned storage medium can be a non-volatile storage medium such as a disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0175] It should be understood that, although the steps in the flowchart of the accompanying drawings are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

[0176] Further references Figure 3 , as a response to the above Figure 2 The present application provides an embodiment of a signature data processing device, and the device embodiment is similar to Figure 2 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0177] like Figure 3 As shown, the signature data processing device 300 described in this embodiment includes: a first acquisition module 301, a first detection module 302, a second detection module 303, a third detection module 304, an extraction module 305 and a comparison module 306. Among them:

[0178] The first acquisition module 301 is used to acquire image data corresponding to the user collected during the signing process;

[0179] A first detection module 302, configured to perform detection processing on the image data based on a preset terminal detection model to obtain a corresponding terminal detection result;

[0180] The second detection module 303 is used to call a preset action detection model to perform signature action detection on the image data based on the terminal detection result to obtain a corresponding signature action detection result;

[0181] The third detection module 304 is used to call a preset trajectory detection model to perform signature trajectory detection on the image data based on the terminal detection result if the signature action detection result matches the preset signature action pattern, so as to obtain a corresponding signature trajectory detection result;

[0182] An extraction module 305 is used to extract the signature from the image data to obtain corresponding signature data if the signature trajectory detection result matches a preset signature trajectory pattern;

[0183] The comparison module 306 is used to compare the signature data based on a preset signature recognition model and pre-stored signature information to obtain a signature comparison result corresponding to the signature data.

[0184] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the signature data processing method in the aforementioned implementation mode, and will not be repeated here.

[0185] In some optional implementations of this embodiment, the second detection module 303 includes:

[0186] a determination submodule, configured to determine a first position of the hand in the image data based on the terminal detection result, and obtain a hand region image corresponding to the first position;

[0187] A first processing submodule, used for preprocessing the hand region image to obtain a corresponding first image;

[0188] A first detection submodule, configured to perform hand key point detection on the first image based on a preset hand key point detection algorithm to obtain corresponding target key points;

[0189] An extraction submodule is used to extract features of the target key points to obtain corresponding target features;

[0190] A first classification submodule, used to classify the target feature based on the action detection model to obtain a corresponding first classification result;

[0191] The first generating submodule is used to generate the signature action detection result based on the first classification result.

[0192] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the signature data processing method in the aforementioned implementation mode, and will not be repeated here.

[0193] In some optional implementations of this embodiment, the third detection module 304 includes:

[0194] A second processing submodule, configured to determine a second position of the signature area in the image data based on the terminal detection result, and obtain a signature image corresponding to the second position;

[0195] A third processing submodule is used to pre-process the signature image to obtain a corresponding second image;

[0196] A second classification submodule, used to classify the second image based on the trajectory detection model to obtain a corresponding second classification result;

[0197] The second generating submodule is used to generate the signature trajectory detection result based on the second classification result.

[0198] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the signature data processing method in the aforementioned implementation mode, and will not be repeated here.

[0199] In some optional implementations of this embodiment, the extraction module 305 includes:

[0200] A first acquisition submodule, used to acquire a signature position in the image data;

[0201] A cropping submodule, used for cropping the image data based on the signature position to obtain a corresponding third image;

[0202] A second detection submodule, configured to perform edge detection processing on the third image based on a preset edge detection algorithm to obtain a corresponding fourth image;

[0203] The third generating submodule is used to generate the signature data based on the fourth image.

[0204] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the signature data processing method in the aforementioned implementation mode, and will not be repeated here.

[0205] In some optional implementations of this embodiment, the comparison module 306 includes:

[0206] A second acquisition submodule is used to acquire a specified format corresponding to the signature information;

[0207] An identification submodule, configured to perform identification processing on the signature data based on the signature identification model to obtain designated information corresponding to the designated form;

[0208] A third acquisition submodule is used to acquire a specified comparison strategy corresponding to the specified form;

[0209] The comparison submodule is used to compare the specified information with the signature information based on the specified comparison strategy to obtain a signature comparison result corresponding to the signature data.

[0210] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the signature data processing method in the aforementioned implementation mode, and will not be repeated here.

[0211] In some optional implementations of this embodiment, the comparison submodule includes:

[0212] A calling unit, configured to call a preset similarity algorithm if the specified form of the signature information is a feature vector form;

[0213] a calculation unit, configured to calculate the similarity between the specified information and the signature information based on the similarity algorithm;

[0214] A judging unit, used to judge whether the similarity is greater than a preset similarity threshold;

[0215] A first generating unit, configured to generate a first comparison result indicating that the signature is a genuine signature;

[0216] The second generating unit is used to generate a second comparison result indicating that the signature is a forged signature if no.

[0217] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the signature data processing method in the aforementioned implementation mode, and will not be repeated here.

[0218] In some optional implementations of this embodiment, the signature data processing device further includes:

[0219] A calling module is used to call a preset initial detection model and obtain a preset engineering transformation strategy;

[0220] A transformation module, used to transform the initial detection model based on the engineering transformation strategy to obtain a corresponding first detection model;

[0221] A second acquisition module is used to acquire pre-collected initial sample data related to terminal detection;

[0222] An enhancement module, used to perform data enhancement processing on the initial sample data to obtain corresponding target sample data;

[0223] A training module, used to train the first detection model based on the target sample data to obtain a second detection model that meets the preset construction requirements;

[0224] A determination module is used to use the second detection model as the terminal detection model.

[0225] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the signature data processing method in the aforementioned implementation mode, and will not be repeated here.

[0226] To solve the above technical problems, the present application also provides a computer device. Figure 4 , Figure 4 This is a basic structural block diagram of the computer device in this embodiment.

[0227] The computer device 4 includes a memory 41, a processor 42, and a network interface 43 that are interconnected through a system bus. It should be noted that the figure only shows a computer device 4 with components 41-43, but it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented instead. Among them, those skilled in the art can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application specific integrated circuits (Application Specific Integrated Circuit, ASIC), programmable gate arrays (Field-Programmable Gate Array, FPGA), digital processors (Digital Signal Processor, DSP), embedded devices, etc.

[0228] The computer device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The computer device may interact with a user through a keyboard, a mouse, a remote controller, a touch pad, or a voice control device.

[0229] The memory 41 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, etc. In some embodiments, the memory 41 can be an internal storage unit of the computer device 4, such as a hard disk or memory of the computer device 4. In other embodiments, the memory 41 can also be an external storage device of the computer device 4, such as a plug-in hard disk equipped on the computer device 4, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (FlashCard), etc. Of course, the memory 41 can also include both the internal storage unit of the computer device 4 and its external storage device. In this embodiment, the memory 41 is generally used to store the operating system and various application software installed on the computer device 4, such as computer-readable instructions of the signature data processing method, etc. In addition, the memory 41 can also be used to temporarily store various types of data that have been output or are to be output.

[0230] The processor 42 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips in some embodiments. The processor 42 is generally used to control the overall operation of the computer device 4. In this embodiment, the processor 42 is used to run computer-readable instructions or process data stored in the memory 41, such as computer-readable instructions for running the signature data processing method.

[0231] The network interface 43 may include a wireless network interface or a wired network interface. The network interface 43 is generally used to establish a communication connection between the computer device 4 and other electronic devices.

[0232] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0233] In an embodiment of the present application, after obtaining the image data corresponding to the user collected during the signing process, the image data is subjected to terminal detection, signature action detection, signature trajectory detection, and signature comparison by combining the use of a terminal detection model, an action detection model, a trajectory detection model, and a signature recognition model to generate a final signature comparison result, thereby effectively improving the comprehensiveness of the signature authenticity verification detection and improving the accuracy of the generated signature comparison result.

[0234] The present application also provides another embodiment, namely, providing a computer-readable storage medium, wherein the computer-readable storage medium stores computer-readable instructions, and the computer-readable instructions can be executed by at least one processor to enable the at least one processor to perform the steps of the signature data processing method as described above.

[0235] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0236] In an embodiment of the present application, after obtaining the image data corresponding to the user collected during the signing process, the image data is subjected to terminal detection, signature action detection, signature trajectory detection, and signature comparison by combining the use of a terminal detection model, an action detection model, a trajectory detection model, and a signature recognition model to generate a final signature comparison result, thereby effectively improving the comprehensiveness of the signature authenticity verification detection and improving the accuracy of the generated signature comparison result.

[0237] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a terminal device (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0238] Obviously, the embodiments described above are only some embodiments of the present application, rather than all embodiments. The preferred embodiments of the present application are given in the accompanying drawings, but they do not limit the patent scope of the present application. The present application can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present application more thorough and comprehensive. Although the present application is described in detail with reference to the aforementioned embodiments, for those skilled in the art, it is still possible to modify the technical solutions recorded in the aforementioned specific implementation methods, or to perform equivalent replacement of some of the technical features therein. Any equivalent structure made using the contents of the specification and drawings of this application, directly or indirectly used in other related technical fields, is similarly within the scope of patent protection of this application.

Claims

1. A method for processing signature data, characterized in that: The steps include: Obtaining image data corresponding to the user collected during the signing process; Performing detection processing on the image data based on a preset terminal detection model to obtain a corresponding terminal detection result; Based on the terminal detection result, calling a preset action detection model to perform signature action detection on the image data to obtain a corresponding signature action detection result; If the signature action detection result matches the preset signature action pattern, based on the terminal detection result, calling a preset trajectory detection model to perform signature trajectory detection on the image data to obtain a corresponding signature trajectory detection result; If the signature trajectory detection result matches the preset signature trajectory pattern, the signature is extracted from the image data to obtain corresponding signature data; The signature data is compared with the pre-stored signature information based on a preset signature recognition model to obtain a signature comparison result corresponding to the signature data.

2. The method for processing signature data according to claim 1, characterized in that: The step of calling a preset action detection model to perform signature action detection on the image data based on the terminal detection result to obtain a corresponding signature action detection result specifically includes: Determine a first position of the hand in the image data based on the terminal detection result, and acquire a hand area image corresponding to the first position; Preprocessing the hand region image to obtain a corresponding first image; Performing hand key point detection on the first image based on a preset hand key point detection algorithm to obtain corresponding target key points; Extracting features of the target key points to obtain corresponding target features; Classify the target feature based on the action detection model to obtain a corresponding first classification result; The signature action detection result is generated based on the first classification result.

3. The method for processing signature data according to claim 1, characterized in that: The step of calling a preset trajectory detection model to perform signature trajectory detection on the image data based on the terminal detection result to obtain a corresponding signature trajectory detection result specifically includes: Determine a second position of the signature area in the image data based on the terminal detection result, and obtain a signature image corresponding to the second position; Preprocessing the signature image to obtain a corresponding second image; Performing classification processing on the second image based on the trajectory detection model to obtain a corresponding second classification result; The signature trajectory detection result is generated based on the second classification result.

4. The method for processing signature data according to claim 1, characterized in that: The step of extracting the signature from the image data to obtain the corresponding signature data specifically includes: Obtaining a signature position in the image data; Cropping the image data based on the signature position to obtain a corresponding third image; Performing edge detection processing on the third image based on a preset edge detection algorithm to obtain a corresponding fourth image; The signature data is generated based on the fourth image.

5. The method for processing signature data according to claim 1, characterized in that: The step of comparing the signature data based on a preset signature recognition model with pre-stored signature information to obtain a signature comparison result corresponding to the signature data specifically includes: Obtaining a specified form corresponding to the signature information; Performing recognition processing on the signature data based on the signature recognition model to obtain designated information corresponding to the designated form; Obtaining a specified comparison strategy corresponding to the specified form; The designated information and the signature information are compared based on the designated comparison strategy to obtain a signature comparison result corresponding to the signature data.

6. The method for processing signature data according to claim 5, characterized in that: The step of comparing the designated information with the signature information based on the designated comparison strategy to obtain a signature comparison result corresponding to the signature data specifically includes: If the specified form of the signature information is a feature vector form, calling a preset similarity algorithm; Calculating the similarity between the specified information and the signature information based on the similarity algorithm; Determining whether the similarity is greater than a preset similarity threshold; If so, the generated signature is the first comparison result of the real signature; If not, a second comparison result is generated indicating that the signature is a forged signature.

7. The method for processing signature data according to claim 1, characterized in that: Before the step of performing detection processing on the image data based on the preset terminal detection model to obtain the corresponding terminal detection result, the method further includes: Call the preset initial detection model and obtain the preset engineering transformation strategy; Based on the engineering transformation strategy, the initial detection model is transformed to obtain a corresponding first detection model; Obtaining pre-collected initial sample data related to terminal detection; Performing data enhancement processing on the initial sample data to obtain corresponding target sample data; Training the first detection model based on the target sample data to obtain a second detection model that meets the preset construction requirements; The second detection model is used as the terminal detection model.

8. A signature data processing device, characterized in that: include: A first acquisition module, used to acquire image data corresponding to the user collected during the signing process; A first detection module, used to perform detection processing on the image data based on a preset terminal detection model to obtain a corresponding terminal detection result; A second detection module is used to call a preset action detection model to perform signature action detection on the image data based on the terminal detection result to obtain a corresponding signature action detection result; A third detection module is used to call a preset trajectory detection model to perform signature trajectory detection on the image data based on the terminal detection result if the signature action detection result matches the preset signature action pattern, so as to obtain a corresponding signature trajectory detection result; An extraction module, configured to extract the signature from the image data to obtain corresponding signature data if the signature trajectory detection result matches a preset signature trajectory pattern; The comparison module is used to compare the signature data based on a preset signature recognition model and pre-stored signature information to obtain a signature comparison result corresponding to the signature data.

9. A computer device, characterized in that: The method comprises a memory and a processor, wherein the memory stores computer-readable instructions, and the processor implements the steps of the signature data processing method as claimed in any one of claims 1 to 7 when executing the computer-readable instructions.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by a processor, the steps of the signature data processing method according to any one of claims 1 to 7 are implemented.

Images