Intelligent review method, device, storage medium and terminal for matching person and vehicle

Abstract

The present invention relates to the field of face recognition, and proposes an intelligent audit method, device, storage medium and terminal for matching people and vehicles, wherein the method comprises: obtaining a facial image of a person to be tested, and determining the area to be tested of the facial image according to a preset image face detection model; determining the facial features of the area to be tested according to a preset face feature extraction model; comparing the facial features with pre-stored facial features, and if the similarity is greater than or equal to a preset threshold, performing a liveness detection on the person to be tested, and if the liveness detection is successful, obtaining the pre-stored license plate number of the person to be tested; confirming the license plate number of the vehicle to be tested according to a preset license plate recognition model; and determining the audit result according to the pre-stored license plate number and the license plate number. The method of the present invention ensures that the audit process is implemented by the driver himself, and prevents cheating during the audit process.

Description

Technical Field

[0001] The present invention relates to the field of face recognition technology, and in particular to an intelligent verification method, device, storage medium and terminal for matching a person with a vehicle. Background Art

[0002] Online car-hailing has become an important way of traveling. During the operation, it is necessary to ensure that the vehicle operated by the driver is consistent with the vehicle bound to the backend to ensure passenger safety and improve user experience. Therefore, regular human-vehicle matching review plays a very important role.

[0003] The manual human-vehicle matching review method has a large number of reviews, low efficiency, poor timeliness, and is difficult to implement. The intelligent review method uses a camera to obtain a panoramic image, identify the human-vehicle associated target in the panoramic image, obtain the face information and vehicle information of the human-vehicle associated target, and judge whether the human-vehicle associated target matches based on the comparison of the face information and vehicle information. This method requires a panoramic image to be obtained through a camera, which has high requirements for the shooting distance and shooting angle, and other personnel need to intervene during the shooting process, and it is impossible for a single person to complete the image shooting; secondly, traditional technology cannot handle the situation of facial information cheating during the driver's photo shooting process, thereby interfering with the investigation of human-vehicle matching. Summary of the invention

[0004] In order to solve at least one of the above technical problems, the present invention proposes a method, device, storage medium and terminal for intelligent review of person-vehicle matching, and the present invention is specifically implemented by the following technical solutions:

[0005] The first aspect of the present invention provides an intelligent review method for matching a person with a vehicle, the method comprising:

[0006] Acquire a facial image of a person to be tested, and determine a region to be tested of the facial image according to a preset image face detection model;

[0007] Determining the facial features of the area to be tested according to a preset facial feature extraction model;

[0008] Compare the facial features with pre-stored facial features, and if the similarity is greater than or equal to a preset threshold, perform liveness detection on the person to be tested, and if the liveness detection is successful, obtain the pre-stored license plate number of the person to be tested;

[0009] According to the preset license plate recognition model, confirm the license plate number of the vehicle to be tested;

[0010] The audit result is determined based on the pre-stored license plate number and the license plate number.

[0011] In some possible implementations, the performing liveness detection on the person to be tested includes:

[0012] Obtain a facial video of the person to be tested, determine the face position of the facial video according to a preset video face detection model, and capture the image to be tested from the face position;

[0013] Determine the facial feature points of the image to be tested according to a preset key point positioning model;

[0014] According to a preset facial action model, facial action detection is performed on the facial video based on the facial feature points;

[0015] If the number of facial actions in the facial video is greater than or equal to a preset number threshold, the detection is successful.

[0016] In some possible implementations, determining the area to be detected of the facial image according to a preset image face detection model includes:

[0017] Identify the key points of the facial features of the facial image according to a preset image face detection model;

[0018] Face alignment is performed according to the key points of the facial features to determine the area to be tested of the facial image.

[0019] In some possible implementations, confirming the license plate number of the vehicle to be tested according to a preset license plate recognition model includes:

[0020] Get the license plate image of the vehicle to be tested;

[0021] According to a preset license plate detection model, determining a license plate area to be detected in the license plate image;

[0022] According to a preset license plate recognition model, the license plate number in the license plate area to be tested is determined.

[0023] In some possible implementations, the method further includes:

[0024] Constructing a face and license plate database, wherein the face and license plate database is used to store user IDs, pre-stored license plate numbers, and pre-stored face images;

[0025] According to the user ID, the pre-stored license plate number and the pre-stored face image are obtained.

[0026] In some possible implementations, comparing the facial features with pre-stored facial features includes:

[0027] Determine the pre-stored facial features of the pre-stored facial images in the facial license plate database according to a preset facial feature extraction model;

[0028] Calculate the similarity between the facial features and preset facial features.

[0029] In some possible implementations, determining the audit result according to the facial features and the license plate number includes:

[0030] According to the license plate number, obtaining the face image in the face license plate database;

[0031] Determining the facial features to be tested of the facial image according to a preset facial feature extraction model;

[0032] Calculating the similarity between the facial features and the facial features to be tested;

[0033] If the similarity is greater than or equal to the preset threshold, the review is successful.

[0034] A second aspect of the present invention provides an intelligent verification device for matching a person with a vehicle, the device comprising:

[0035] A face detection module is used to obtain a facial image of a person to be tested and determine a region to be tested of the facial image according to a preset image face detection model;

[0036] A facial feature extraction module, used to determine the facial features of the area to be tested according to a preset facial feature extraction model;

[0037] A liveness detection module is used to compare the facial features with pre-stored facial features. If the similarity is greater than or equal to a preset threshold, liveness detection is performed on the person to be tested. If the liveness detection is successful, the pre-stored license plate number of the person to be tested is obtained;

[0038] The license plate number recognition module is used to confirm the license plate number of the vehicle to be tested according to the preset license plate recognition model;

[0039] The result acquisition module is used to determine the audit result based on the pre-stored license plate number and the license plate number.

[0040] In some possible implementations, the living body detection module includes:

[0041] A face position determination unit, used to obtain a face video of a person to be tested, determine the face position of the face video according to a preset video face detection model, and capture a test image from the face position;

[0042] A facial feature point determination unit, used to determine the facial feature points of the image to be tested according to a preset key point positioning model;

[0043] A facial action detection unit, configured to perform facial action detection on the facial video based on the facial feature points according to a preset facial action model;

[0044] The detection result determination unit is used to determine that the detection is successful if the number of facial actions in the facial video is greater than or equal to a preset number threshold.

[0045] The present invention also provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned intelligent review method for matching a person with a vehicle is implemented.

[0046] The present invention also provides a terminal, comprising one or more processors and a memory. The memory is coupled to the processor and is used to store one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the above-mentioned intelligent vehicle matching review method.

[0047] By adopting the above technical solution, the intelligent human-vehicle matching review method and device of the present invention have the following beneficial effects:

[0048] By quickly detecting the facial features of the person under test, we can quickly determine whether the person under test is an approved driver, and then perform liveness detection on the person under test's face to confirm that the test was conducted by the person under test himself, ensuring that the person under test is a qualified driver and that the vehicle operated by the person under test is consistent with the vehicle bound to the background, avoiding cheating during the intelligent review process. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0050] Figure 1 A flow chart of an intelligent review method for matching a person with a vehicle provided by an embodiment of the present invention;

[0051] Figure 2 A structural diagram of an intelligent vehicle matching review device provided by an embodiment of the present invention;

[0052] Figure 3 A structural diagram of a living body detection module provided in an embodiment of the present invention;

[0053] Figure 4 A schematic diagram of the structure of a computer terminal device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0054] The technical scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is only a part of the embodiment of the present invention, not all of the embodiments. Based on the embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0055] In several embodiments provided in this application, the described system embodiments are only schematic, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation, for example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, indirect coupling or communication connection of modules or units, which may be electrical or other forms.

[0056] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed on multiple network modules. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0057] In addition, each functional unit in each embodiment of the present invention may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0058] The embodiment of the present invention provides an intelligent review method for matching a person with a vehicle, such as Figure 1 As shown, the method includes:

[0059] S101, obtaining a facial image of a person to be tested, and determining a region to be tested of the facial image according to a preset image face detection model;

[0060] The preset image face detection model includes a multi-task cascaded convolutional network (MTCNN) model. To build the preset image face detection model, it is necessary to input a face image and use the face detection and key point positioning algorithm to train the model. The person to be tested logs in to the vehicle detection APP, uses the facial recognition module in the APP to take a selfie of the face image, uploads the face image to the preset image face detection model, locates the facial features of the face image, and determines the face area to be tested in the face image.

[0061] S102, determining the facial features of the area to be tested according to a preset facial feature extraction model;

[0062] The preset face feature extraction model needs to be pre-trained using a residual network (ResNet). The trained preset face feature extraction model can accurately extract face features. The area to be tested is input into the preset face feature extraction model to obtain face features of the face in the area to be tested.

[0063] S103, comparing the facial features with pre-stored facial features, and if the similarity is greater than or equal to a preset threshold, performing a liveness detection on the person to be tested, and if the liveness detection is successful, obtaining the pre-stored license plate number of the person to be tested;

[0064] It is necessary to collect facial images of all drivers in advance, and use the predicted facial feature extraction model to obtain the preset facial features of each driver's facial image, and store the preset facial features in the database. In this way, when obtaining the facial features of the face of the person to be tested, it is necessary to traverse the preset facial features from the database and calculate the similarity with the facial features of the face of the person to be tested. If the similarity is less than the preset threshold, such as the similarity is less than 72%, it means that there is no record of the person to be tested, so the intelligent review of person-vehicle matching fails; if the similarity is greater than or equal to the preset threshold, such as the similarity is greater than or equal to 72%, it means that there is a record of the person to be tested, and the liveness detection of the person to be tested is continued to confirm whether the review is conducted by the person to be tested himself. The person to be tested takes a selfie with a mobile phone and aligns the face and the license plate number area. If the liveness detection fails, the intelligent review of person-vehicle matching fails; if the liveness detection is successful, the pre-stored license plate number is obtained according to the account logged in by the person to be tested;

[0065] S104, confirming the license plate number of the vehicle to be tested according to a preset license plate recognition model;

[0066] The license plate recognition model is trained using a feature extraction neural network (Convolutional Recurrent Neural Network, CRNN) to build a preset license plate recognition model. According to the preset license plate recognition model, a screenshot is taken from the video to obtain a picture with the license plate number area and the license plate number is extracted.

[0067] S105. Determine the audit result according to the pre-stored license plate number and the license plate number.

[0068] Use the account logged in by the person to be tested to obtain the pre-stored license plate number, that is, the registered license plate number, and compare the pre-stored license plate number with the license plate number recognized from the picture. If they are the same, the review is successful. If they are different, the review fails.

[0069] In one embodiment, the performing liveness detection on the person to be tested includes:

[0070] Obtain a facial video of the person to be tested, determine the face position of the facial video according to a preset video face detection model, and capture the image to be tested from the face position;

[0071] Determine the facial feature points of the image to be tested according to a preset key point positioning model;

[0072] According to a preset facial action model, facial action detection is performed on the facial video based on the facial feature points;

[0073] If the number of facial actions in the facial video is greater than or equal to a preset number threshold, the detection is successful.

[0074] A face detection model is trained based on a cascaded convolutional neural network (Cascaded RCNN) to obtain a preset face detection model; a preset key point positioning model is trained using a face key point positioning algorithm, and 81 facial key points on the face can be obtained through the key point positioning model; the person to be tested uses a mobile phone to aim at his or her face to shoot a video image to obtain a facial video, and then uses a preset face detection model to determine the face position in the facial video, and inputs the face position into the preset key point positioning model to obtain a face image marked with feature points, and then based on the 81 feature points of the face, using a preset facial action model, guides the person to be tested to complete live actions such as blinking, nodding, shaking the head, and opening the mouth. If the number of facial actions of the person to be tested in the facial video is less than a preset number threshold, the liveness detection fails, and the intelligent review of the person-vehicle matching fails; if the number of facial actions of the person to be tested is greater than or equal to the preset number threshold, the liveness detection is successful; through the liveness detection, it is determined that the test is conducted by the person to be tested himself or herself, rather than cheating by detecting photos or the like.

[0075] In one embodiment, determining the area to be detected of the facial image according to a preset image face detection model includes:

[0076] Identify the key points of the facial features of the facial image according to a preset image face detection model;

[0077] Face alignment is performed according to the key points of the facial features to determine the area to be tested of the facial image.

[0078] The image face detection model is pre-trained using the MTCNN model, and the facial image of the person to be tested is input to locate the key points of the facial features. If the person to be tested does not face the camera when taking the photo, the key points of the facial features are tilted. At this time, the image face detection model needs to align the face according to the key points of the facial features to obtain a clear facial image, and then determine the area to be tested in the facial image according to the key points of the facial features.

[0079] In one embodiment, the step of confirming the license plate number of the vehicle to be tested according to a preset license plate recognition model includes:

[0080] Get the license plate image of the vehicle to be tested;

[0081] According to a preset license plate detection model, determining a license plate area to be detected in the license plate image;

[0082] According to a preset license plate recognition model, the license plate number in the license plate area to be tested is determined.

[0083] A license plate detection model is trained using a text detection algorithm Differentiable Binarization (DBnet) network to build a preset license plate detection model; a license plate recognition model is trained using a feature extraction neural network (CRNN) to build a preset license plate recognition model; when the liveness detection of the person to be tested passes, a screenshot is taken from the video containing the face and license plate number of the person to be tested, the license plate image of the vehicle to be tested is obtained, the license plate image is input into the preset license plate detection model, the license plate area to be tested in the license plate image is located, the license plate area to be tested is input into the preset license plate recognition model, and the license plate number of the vehicle to be tested is identified.

[0084] In one embodiment, the method further comprises:

[0085] Constructing a face and license plate database, wherein the face and license plate database is used to store user IDs, pre-stored license plate numbers, and pre-stored face images;

[0086] According to the user ID, the pre-stored license plate number and the pre-stored face image are obtained.

[0087] Build a face and license plate database, and establish a record of the driver's user number, face image and license plate number in one-to-one correspondence; the face and license plate database includes the user number, pre-stored license plate number and pre-stored face image; when the user logs in to the human-vehicle detection APP, obtain the exclusive user number, and obtain the corresponding pre-stored face image and pre-stored license plate number from the face and license plate database according to the user number;

[0088] In one embodiment, comparing the facial features with pre-stored facial features includes:

[0089] Determine the pre-stored facial features of the face image in the face license plate database according to a preset facial feature extraction model;

[0090] Calculate the similarity between the facial features and preset facial features.

[0091] By inputting the driver's photo in the database into the preset facial feature extraction model, the pre-stored facial features of the pre-stored face picture are obtained, and the facial features of the person to be tested are obtained through the preset image face detection model and the preset facial feature extraction model, and then the similarity between the facial features of the person to be tested and the pre-stored facial features is calculated. If the similarity is less than the preset threshold, such as the similarity is less than 72%, it means that there is no record information of the person to be tested, and the review fails. If the similarity is greater than or equal to the preset threshold, such as the preset threshold is 72%, it means that there is record information of the person to be tested, and the facial liveness detection continues. Initially, the facial image is used to detect whether the facial features of the person to be tested are the facial features of the same person as the pre-stored facial features in the login account. If they are not the facial features of the same person, subsequent detection can be avoided, saving computing resources.

[0092] In one embodiment, determining the audit result according to the facial features and the license plate number includes:

[0093] According to the license plate number, obtaining the face image in the face license plate database;

[0094] Determining the facial features to be tested of the facial image according to a preset facial feature extraction model;

[0095] Calculating the similarity between the facial features and the facial features to be tested;

[0096] If the similarity is greater than or equal to the preset threshold, the review is successful.

[0097] Specifically, using the license plate number, the facial image corresponding to the license plate number is queried from the database, the pre-stored facial features corresponding to the facial image are obtained, and the similarity between the facial features and the pre-stored facial features is calculated. If the similarity is less than a preset threshold, such as the similarity is less than 72%, it means that the license plate number does not match the facial features of the person to be tested, that is, the vehicle to be tested is not the vehicle registered by the driver to be tested in the record, and therefore, the intelligent review of the person-vehicle matching fails; if the similarity is greater than or equal to the preset threshold, such as the similarity is greater than or equal to 72%, the intelligent review of the person-vehicle matching is successful, that is, the vehicle to be tested is the vehicle registered by the person to be tested in the system.

[0098] See also Figure 2The second aspect of the present invention provides an intelligent vehicle matching review device, the device comprising:

[0099] The face detection module 10 is used to obtain a facial image of a person to be tested and determine a region to be tested of the facial image according to a preset image face detection model;

[0100] A face feature extraction module 20, used to determine the face features of the area to be tested according to a preset face feature extraction model;

[0101] The liveness detection module 30 is used to compare the facial features with the pre-stored facial features. If the similarity is greater than or equal to a preset threshold, the liveness detection is performed on the person to be tested. If the liveness detection is successful, the pre-stored license plate number of the person to be tested is obtained.

[0102] The license plate number recognition module 40 is used to confirm the license plate number of the vehicle to be tested according to a preset license plate recognition model;

[0103] The result acquisition module 50 is used to determine the audit result according to the pre-stored license plate number and the license plate number.

[0104] See also Figure 3 In one embodiment, the living body detection module includes:

[0105] The face position determination unit 31 is used to obtain a face video of the person to be tested, determine the face position of the face video according to a preset video face detection model, and capture the image to be tested from the face position;

[0106] A facial feature point determination unit 32, used to determine the facial feature points of the image to be tested according to a preset key point positioning model;

[0107] A facial action detection unit 33, configured to perform facial action detection on the facial video based on the facial feature points according to a preset facial action model;

[0108] The detection result determination unit 34 is configured to determine that the detection is successful if the number of the facial action in the facial video is greater than or equal to a preset number threshold.

[0109] For the specific limitations of the intelligent review device for matching people and vehicles, please refer to the limitations mentioned above and will not be repeated here.

[0110] Each module in the above-mentioned intelligent vehicle matching review device can be implemented in whole or in part by software, hardware, or a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0111] See also Figure 4 An embodiment of the present invention provides a terminal, comprising one or more processors and a memory. The memory is coupled to the processor and is used to store one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the intelligent review method for matching a person with a vehicle in any of the above embodiments.

[0112] The processor is used to control the overall operation of the computer terminal device to complete all or part of the steps of the above-mentioned intelligent vehicle matching audit method. The memory is used to store various types of data to support the operation of the computer terminal device, and these data may include, for example, instructions for any application or method used to operate on the computer terminal device, as well as application-related data. The memory can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read-Only Memory, referred to as EPROM), programmable read-only memory (Programmable Read-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic memory, flash memory, disk or optical disk.

[0113] In an exemplary embodiment, the computer terminal device can be implemented by one or more application specific integrated circuits (Application Specific Integrated Circuit, ASC), digital signal processors (Digital Signal Processor, DSP), digital signal processing devices (Digital Signal Processing Device, DSPD), programmable logic devices (Programmable Logic Device, PLD), field programmable gate arrays (Field Programmable Gate Array, FPGA), controllers, microcontrollers, microprocessors or other electronic components to execute the above-mentioned intelligent human-vehicle matching review method and achieve the same technical effect as the above-mentioned method.

[0114] In another exemplary embodiment, a storage medium including program instructions is also provided, and when the program instructions are executed by a processor, the steps of the intelligent review method for matching people and vehicles in any of the above embodiments are implemented. For example, the storage medium may be the above-mentioned memory including program instructions, and the above-mentioned program instructions may be executed by a processor of a terminal to complete the above-mentioned intelligent review method for matching people and vehicles, and achieve the same technical effect as the above-mentioned method.

[0115] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An intelligent human-vehicle matching review method, characterized in that: The method comprises: Acquire a facial image of a person to be tested, and determine a test area of ​​the facial image according to a preset image face detection model, wherein the facial image is obtained by the person to be tested photographing his or her own face; Determining the facial features of the area to be tested according to a preset facial feature extraction model; Compare the facial features with the pre-stored facial features. If the similarity is less than a preset threshold, determine that the audit has failed; if the similarity is greater than or equal to the preset threshold, perform a liveness detection on the person to be tested. If the liveness detection is successful, obtain the pre-stored license plate number of the person to be tested, and the pre-stored facial features are obtained after the person to be tested logs into the current application; Obtaining a license plate image of the vehicle to be tested, wherein the license plate image is obtained by taking a screenshot of a facial video taken by the person to be tested, including the face and the license plate number; According to a preset license plate detection model, determining a license plate area to be detected in the license plate image; According to a preset license plate recognition model, determining the license plate number in the license plate area to be tested; The audit result is determined based on the pre-stored license plate number and the license plate number.

2. The method according to claim 1, characterized in that The performing liveness detection on the person to be tested comprises: Obtain a facial video of the person to be tested, determine the face position of the facial video according to a preset video face detection model, and capture the image to be tested from the face position; Determine the facial feature points of the image to be tested according to a preset key point positioning model; According to a preset facial action model, facial action detection is performed on the facial video based on the facial feature points; If the number of facial actions in the facial video is greater than or equal to a preset number threshold, the detection is successful.

3. The method according to claim 1, characterized in that The step of determining the area to be detected of the facial image according to a preset image face detection model comprises: Identify the key points of the facial features of the facial image according to a preset image face detection model; Face alignment is performed according to the key points of the facial features to determine the area to be tested of the facial image.

4. The method according to claim 1, characterized in that: The method further comprises: Constructing a face and license plate database, wherein the face and license plate database is used to store user IDs, pre-stored license plate numbers, and pre-stored face images; According to the user ID, the pre-stored license plate number and the pre-stored face image are obtained.

5. The method according to claim 4, characterized in that The comparing the facial features with pre-stored facial features includes: Determine the pre-stored facial features of the pre-stored facial images in the facial license plate database according to a preset facial feature extraction model; Calculate the similarity between the facial features and preset facial features.

6. An intelligent verification device for matching a person with a vehicle, characterized in that: include: A face detection module is used to obtain a facial image of a person to be tested, and determine a region to be tested of the facial image according to a preset image face detection model, wherein the facial image is obtained by the person to be tested photographing his or her own face; A facial feature extraction module, used to determine the facial features of the area to be tested according to a preset facial feature extraction model; A liveness detection module is used to compare the facial features with pre-stored facial features. If the similarity is less than a preset threshold, it is determined that the audit has failed; if the similarity is greater than or equal to the preset threshold, a liveness detection is performed on the person to be tested. If the liveness detection is successful, the pre-stored license plate number of the person to be tested is obtained. The pre-stored facial features are obtained after the person to be tested logs into the current application; A license plate image acquisition module is used to acquire a license plate image of the vehicle to be tested, wherein the license plate image is obtained by taking a screenshot of a facial video including the face and the license plate number taken by the person to be tested; A license plate area determination module to be tested, used to determine the license plate area to be tested in the license plate image according to a preset license plate detection model; A license plate number recognition module for determining the license plate number in the license plate area to be tested according to a preset license plate recognition model; The result acquisition module is used to determine the audit result based on the pre-stored license plate number and the license plate number.

7. The device according to claim 6, characterized in that The living body detection module comprises: A face position determination unit, used to obtain a face video of a person to be tested, determine the face position of the face video according to a preset video face detection model, and capture a test image from the face position; A facial feature point determination unit, used to determine the facial feature points of the image to be tested according to a preset key point positioning model; A facial action detection unit, configured to perform facial action detection on the facial video based on the facial feature points according to a preset facial action model; The detection result determination unit is used to determine that the detection is successful if the number of facial actions in the facial video is greater than or equal to a preset number threshold.

8. A storage medium, characterized in that: The storage medium stores instructions, and when the instructions are executed by the processor, the steps of the method according to any one of claims 1 to 5 are implemented.

9. A terminal, characterized in that: The method comprises a memory and a processor, wherein the memory stores instructions, and the processor loads the instructions to execute the steps of the method according to any one of claims 1 to 5.

Images