Method, device, computer-readable medium, and electronic device for identifying authenticity of certificate

Abstract

The embodiments of the present application provide a method, device, computer-readable medium, and electronic device for identifying the authenticity of a certificate. The method for identifying the authenticity of a certificate includes: detecting dynamic anti-counterfeiting points and static anti-counterfeiting points contained in multiple certificate images of a target certificate, wherein the multiple certificate images are obtained by capturing images of the target certificate at different angles; extracting image feature information of the locations of static anti-counterfeiting points to obtain static anti-counterfeiting point features; extracting image feature information of the locations of dynamic anti-counterfeiting points and feature information of changes in dynamic anti-counterfeiting points between multiple certificate images to obtain dynamic anti-counterfeiting point features; identifying the authenticity results corresponding to each static anti-counterfeiting point based on the static anti-counterfeiting point features, and identifying the authenticity results corresponding to each dynamic anti-counterfeiting point based on the dynamic anti-counterfeiting point features; and determining the authenticity of the target certificate based on the authenticity results corresponding to each static anti-counterfeiting point and the authenticity results corresponding to each dynamic anti-counterfeiting point. The technical solution of the present application can accurately identify the authenticity of a certificate.

Description

Technical Field

[0001] The present application relates to the field of computer and communication technology, and more specifically, to a method and device for identifying the authenticity of a certificate, a computer-readable medium, and an electronic device. Background Art

[0002] In daily life, people often upload images of their IDs (such as ID cards and driver's licenses) online for identity verification. However, during identity verification, there is a risk that the uploaded ID images may be maliciously forged, thus requiring verification of the ID's authenticity. However, methods for verifying ID authenticity proposed in related technologies suffer from low accuracy. Summary of the Invention

[0003] The embodiments of the present application provide a method, apparatus, computer-readable medium, and electronic device for identifying the authenticity of a certificate, thereby accurately identifying the authenticity of a certificate to at least a certain extent.

[0004] Other features and advantages of the present application will become apparent from the following detailed description, or may be learned in part by practice of the present application.

[0005] According to one aspect of an embodiment of the present application, a method for identifying the authenticity of a certificate is provided, comprising: detecting dynamic anti-counterfeiting points and static anti-counterfeiting points contained in multiple certificate images of a target certificate, the multiple certificate images being obtained by capturing images of the target certificate at different angles; extracting image feature information of the positions of the static anti-counterfeiting points to obtain static anti-counterfeiting point features, extracting image feature information of the positions of the dynamic anti-counterfeiting points and feature information of changes in the dynamic anti-counterfeiting points between multiple certificate images to obtain dynamic anti-counterfeiting point features; identifying the authenticity results corresponding to each static anti-counterfeiting point based on the static anti-counterfeiting point features, and identifying the authenticity results corresponding to each dynamic anti-counterfeiting point based on the dynamic anti-counterfeiting point features; determining the authenticity of the target certificate based on the authenticity results corresponding to each static anti-counterfeiting point and the authenticity results corresponding to each dynamic anti-counterfeiting point.

[0006] According to one aspect of an embodiment of the present application, a device for identifying the authenticity of a certificate is provided, comprising: an anti-counterfeiting point detection unit, configured to detect dynamic anti-counterfeiting points and static anti-counterfeiting points contained in multiple certificate images of a target certificate, wherein the multiple certificate images are obtained by capturing images of the target certificate at different angles; an extraction unit, configured to extract image feature information of the positions of the static anti-counterfeiting points to obtain static anti-counterfeiting point features, extract image feature information of the positions of the dynamic anti-counterfeiting points and feature information of changes in the dynamic anti-counterfeiting points between multiple certificate images to obtain dynamic anti-counterfeiting point features; a processing unit, configured to identify the authenticity results corresponding to each static anti-counterfeiting point based on the static anti-counterfeiting point features, and identify the authenticity results corresponding to each dynamic anti-counterfeiting point based on the dynamic anti-counterfeiting point features; a determination unit, configured to determine the authenticity of the target certificate based on the authenticity results corresponding to each static anti-counterfeiting point and the authenticity results corresponding to each dynamic anti-counterfeiting point.

[0007] In some embodiments of the present application, based on the aforementioned scheme, the device for identifying authenticity of a certificate further includes: an acquisition unit, a position detection unit and an acquisition unit; wherein, the acquisition unit is configured to acquire a video stream containing the target certificate from different angles; the position detection unit is configured to detect the position of the target certificate contained in the video image frame in the video stream; the acquisition unit is further configured to: if the position detection unit detects that the position of the target certificate does not meet the set conditions, then re-acquire the video stream containing the target certificate until the position of the target certificate contained in the acquired video image frame meets the set conditions; the acquisition unit is configured to: if it is detected that the position of the target certificate meets the set conditions, then acquire the multiple certificate images from the video image frames contained in the video stream.

[0008] In some embodiments of the present application, based on the aforementioned scheme, the position detection unit is configured to: downsample the video image frame through a plurality of sequentially connected convolution blocks, wherein the first convolution block among the plurality of convolution blocks is used to downsample the video image frame, and the i+1th convolution block among the plurality of convolution blocks is used to downsample the output feature map of the i-th convolution block, where i is greater than 0; upsample the corresponding feature map of the i+1th convolution block among the plurality of convolution blocks in sequence, and merge the result of the upsampling process with the output feature map of the i-th convolution block, and use the merged result as the corresponding feature map of the i-th convolution block, the upsampling process and the downsampling process have the same sampling scale; and identify the position of the target document contained in the video image frame according to the corresponding feature map of the first convolution block among the plurality of convolution blocks.

[0009] In some embodiments of the present application, based on the aforementioned scheme, the document authenticity identification device further includes: an angle detection unit, configured to detect the flip angle of the target document contained in the video image frame in the video stream; the acquisition unit is further configured to: if the detected flip angle of the target document does not meet the set conditions, re-acquire the video stream containing the target document until the flip angle of the target document contained in the acquired video image frame meets the set conditions.

[0010] In some embodiments of the present application, based on the aforementioned scheme, the angle detection unit is configured to: perform target document detection in the video image frame contained in the video stream to identify the document detection frame containing the target document in the video image frame; and determine the flip angle of the target document contained in the video image frame based on the side length ratio of the document detection frame identified in the video image frame.

[0011] In some embodiments of the present application, based on the aforementioned scheme, the extraction unit extracts the image feature information of the location of the dynamic anti-counterfeiting point and the change feature information of the dynamic anti-counterfeiting point between multiple document images to obtain the dynamic anti-counterfeiting point features, and the processing unit identifies the authenticity results corresponding to each dynamic anti-counterfeiting point based on the dynamic anti-counterfeiting point features, including: inputting multiple document image frames containing the dynamic anti-counterfeiting point into a three-dimensional convolutional network to extract a multidimensional feature map of the dynamic anti-counterfeiting point through the three-dimensional convolutional network; converting the multidimensional feature map into a one-dimensional feature map through a downsampling unit in the three-dimensional convolutional network, and outputting the authenticity results corresponding to the dynamic anti-counterfeiting point through a fully connected layer in the three-dimensional convolutional network.

[0012] In some embodiments of the present application, based on the aforementioned scheme, the processing unit is configured to: determine the authenticity result corresponding to each static anti-counterfeiting point according to the first characteristic value interval of the static anti-counterfeiting point characteristics of each static anti-counterfeiting point, and the authenticity result associated with the first characteristic value interval; determine the authenticity result corresponding to each dynamic anti-counterfeiting point according to the second characteristic value interval of the dynamic anti-counterfeiting point characteristics of each dynamic anti-counterfeiting point, and the authenticity result associated with the second characteristic value interval.

[0013] In some embodiments of the present application, based on the aforementioned scheme, the multiple document images include: specified video image frames extracted from a video stream containing the target document, the specified video image frames including any of the following: each video image frame in the video stream, one or more video image frames extracted from the video stream at a set interval, and a video image frame in which the target document is in a horizontal position; the extraction unit is configured to extract image feature information of the location of the static anti-counterfeiting point in the specified video image frame.

[0014] In some embodiments of the present application, based on the aforementioned scheme, the processing unit is configured to: identify the authenticity of the static anti-counterfeiting points contained in each designated video image frame based on the static anti-counterfeiting point features extracted from each designated video image frame; determine the authenticity of the same static anti-counterfeiting point in each designated video image frame based on the authenticity of the static anti-counterfeiting points contained in each designated video image frame; calculate the authenticity of each static anti-counterfeiting point on the target document based on the authenticity of the same static anti-counterfeiting point in each designated video image frame.

[0015] In some embodiments of the present application, based on the aforementioned scheme, the multiple document images include: at least one group of video image frames extracted from a video stream containing the target document; the extraction unit is configured to: extract image feature information of the location of the dynamic anti-counterfeiting point and change feature information of the dynamic anti-counterfeiting point from the at least one group of video image frames.

[0016] In some embodiments of the present application, based on the aforementioned scheme, the processing unit is configured as follows: if multiple groups of video image frames are extracted from the video stream containing the target certificate, then based on the image feature information of the position of the dynamic anti-counterfeiting point extracted from each group of video image frames and the change feature information of the dynamic anti-counterfeiting point, the authenticity of the dynamic anti-counterfeiting point contained in each group of video image frames is identified; based on the authenticity of the dynamic anti-counterfeiting point contained in each group of video image frames, the authenticity of the same dynamic anti-counterfeiting point in each group of video image frames is determined; based on the authenticity of the same dynamic anti-counterfeiting point in each group of video image frames, the authenticity of each dynamic anti-counterfeiting point on the target certificate is calculated.

[0017] In some embodiments of the present application, based on the aforementioned scheme, the authenticity result includes authenticity confidence; the determination unit is configured to: weight the authenticity confidence corresponding to each static anti-counterfeiting point and the authenticity confidence corresponding to each dynamic anti-counterfeiting point according to the weight of each static anti-counterfeiting point and the weight of each dynamic anti-counterfeiting point to obtain a comprehensive authenticity confidence value; determine the authenticity of the target certificate based on the comprehensive authenticity confidence value.

[0018] According to one aspect of an embodiment of the present application, a computer-readable medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the method for identifying the authenticity of a certificate as described in the above embodiment is implemented.

[0019] According to one aspect of an embodiment of the present application, an electronic device is provided, comprising: one or more processors; a storage device for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the document authenticity identification method as described in the above embodiments.

[0020] According to one aspect of an embodiment of the present application, a computer program product or computer program is provided. The computer program product or computer program includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the document authenticity identification method provided in the various optional embodiments described above.

[0021] In the technical solutions provided in some embodiments of the present application, by detecting the dynamic anti-counterfeiting points and static anti-counterfeiting points contained in multiple document images of the target document, it is possible to extract the image feature information of the location of the static anti-counterfeiting points to obtain the static anti-counterfeiting point features, extract the image feature information of the location of the dynamic anti-counterfeiting points and the change feature information of the dynamic anti-counterfeiting points between multiple document images to obtain the dynamic anti-counterfeiting point features, and identify the authenticity results corresponding to each static anti-counterfeiting point based on the static anti-counterfeiting point features, and identify the authenticity results corresponding to each dynamic anti-counterfeiting point based on the dynamic anti-counterfeiting point features, and then determine the authenticity of the target document according to the authenticity results corresponding to each static anti-counterfeiting point and the authenticity results corresponding to each dynamic anti-counterfeiting point. It can be seen that the technical solution of the embodiment of the present application can identify the authenticity of the certificate based on the dynamic anti-counterfeiting points and static anti-counterfeiting points on the certificate image. For the dynamic anti-counterfeiting points, not only the image feature information of the location of the dynamic anti-counterfeiting points is considered, but also the changing feature information of the dynamic anti-counterfeiting points between multiple certificate images is considered, thereby making it possible to identify the authenticity of the target certificate based on the authenticity results corresponding to each static anti-counterfeiting point and the authenticity results corresponding to each dynamic anti-counterfeiting point, thereby improving the accuracy of certificate authenticity identification.

[0022] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The accompanying drawings are incorporated into and constitute a part of the specification, illustrating embodiments consistent with the present application and, together with the specification, explaining the principles of the present application. Obviously, the drawings described below are only some embodiments of the present application, and those skilled in the art can derive other drawings based on these drawings without inventive effort. In the drawings:

[0024] Figure 1 A schematic diagram showing an exemplary system architecture to which the technical solutions of the embodiments of the present application can be applied;

[0025] Figure 2 A flow chart showing a method for identifying the authenticity of a certificate according to an embodiment of the present application is shown;

[0026] Figure 3A schematic diagram of a scenario in which multiple document images are obtained by capturing images of a target document at different angles according to an embodiment of the present application is shown;

[0027] Figure 4 A schematic diagram of a scenario in which multiple document images are obtained by capturing images of a target document at different angles according to an embodiment of the present application is shown;

[0028] Figure 5 A flowchart of detecting the position of a target document based on image segmentation according to an embodiment of the present application is shown;

[0029] Figure 6 A schematic diagram showing a three-dimensional convolutional network identifying authenticity results corresponding to dynamic anti-counterfeiting points according to an embodiment of the present application is shown;

[0030] Figure 7 A flow chart showing a method for identifying the authenticity of a certificate according to an embodiment of the present application is shown;

[0031] Figure 8 A block diagram of a device for identifying authenticity of a certificate according to an embodiment of the present application is shown;

[0032] Figure 9 A schematic diagram of the structure of a computer system suitable for implementing an electronic device according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0033] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this application will be thorough and complete and will fully convey the concepts of the example embodiments to those skilled in the art.

[0034] In addition, described feature, structure or characteristic can be combined in one or more embodiments in any suitable manner.In the following description, many specific details are provided so as to provide a full understanding of the embodiments of the present application. However, it will be appreciated by those skilled in the art that the technical scheme of the present application can be put into practice without one or more of the specific details, or other methods, components, devices, steps etc. can be adopted. In other cases, known methods, devices, implementations or operations are not shown or described in detail to avoid blurring the various aspects of the application.

[0035] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically separate entities. That is, these functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0036] The flowcharts shown in the accompanying drawings are for illustrative purposes only and do not necessarily include all contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may be decomposed, while others may be combined or partially combined. Therefore, the actual execution order may vary depending on the actual situation.

[0037] It should be noted that the term "plurality" used in this document refers to two or more. "And / or" describes a relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A alone, A and B together, or B alone. The character " / " generally indicates an "or" relationship between the associated objects.

[0038] IDs are certificates and documents used to prove identity, experience, and other aspects of life. There are many different types of IDs, such as contactless IC IDs. These contain a built-in chip that provides storage space for basic information such as name, gender, date of birth, and address, as well as a photo. Regardless of the type of ID, it typically includes an anti-counterfeiting mark to facilitate authenticity verification.

[0039] To verify the authenticity of a certificate, the information read from the chip (such as an IC contactless smart certificate) is usually compared with the information derived from the identity data database for consistency, or the facial photo read from the chip and the photo printed on the surface of the certificate are compared for similarity to determine the authenticity of the certificate.

[0040] However, this authenticity verification solution relies on dedicated chip reading tools to read the information, and the verification process also requires manual comparison. The reliability of the entire verification process is closely related to the staff's experience and subjective recognition ability. Everyone's subjective recognition ability is different and has different focuses. The comparison process not only lacks scientific basis, but also lacks unified standards and is easily affected by various factors.

[0041] Based on this, the embodiment of the present application proposes a new method for document authenticity identification, which can identify the authenticity of the document based on the dynamic anti-counterfeiting points and static anti-counterfeiting points on the document image. For the static anti-counterfeiting points, the image feature information of the location of the static anti-counterfeiting points is taken into account. For the dynamic anti-counterfeiting points, not only the image feature information of the location of the dynamic anti-counterfeiting points is taken into account, but also the changing feature information of the dynamic anti-counterfeiting points between multiple document images is taken into account. The authenticity results corresponding to each static anti-counterfeiting point and the authenticity results corresponding to each dynamic anti-counterfeiting point can be determined first, and then the authenticity of the target document can be identified according to the authenticity results corresponding to each static anti-counterfeiting point and the authenticity results corresponding to each dynamic anti-counterfeiting point, thereby improving the accuracy of document authenticity identification.

[0042] Specifically, in a system architecture of this application, if Figure 1 As shown, the system architecture 100 may include a terminal 101 (the terminal 101 may be Figure 1 , or a tablet, laptop, desktop computer, etc.), network 102, and server 103. Network 102 is a medium used to provide a communication link between terminal 101 and server 103. Network 102 may include, but is not limited to, wireless networks and wired networks. Wired networks include, but are not limited to, at least one of the following: a wide area network, a metropolitan area network, and a local area network. Wireless networks include, but are not limited to, at least one of the following: Bluetooth, Wi-Fi, near field communication (NFC), cellular mobile communication networks, etc. A user can use terminal 101 to interact with server 103 via network 102 to receive or send messages, etc.

[0043] It should be understood that Figure 1 The number of terminals 101, networks 102, and servers 103 in the embodiment is merely illustrative. Depending on the implementation requirements, there may be any number of terminals 101, networks 102, and servers 103. For example, the server 103 may be a server cluster consisting of multiple servers.

[0044] In one embodiment of the present application, the terminal 101 can capture images of a target document 104 that needs to be authenticated at different angles, and then send the multiple document images captured at different angles to the server 103 via the network 102. After receiving the multiple document images, the server 103 detects the static anti-counterfeiting points and dynamic anti-counterfeiting points contained in each document image, and then extracts image feature information at the location of the static anti-counterfeiting points to obtain static anti-counterfeiting point features, and extracts image feature information at the location of the dynamic anti-counterfeiting points and feature information on the changes in the dynamic anti-counterfeiting points between the multiple document images to obtain dynamic anti-counterfeiting point features. After obtaining the static anti-counterfeiting point features and the static anti-counterfeiting point features, the authenticity results corresponding to each static anti-counterfeiting point can be identified based on the static anti-counterfeiting point features, and the authenticity results corresponding to each dynamic anti-counterfeiting point can be identified based on the dynamic anti-counterfeiting point features. Then, the authenticity of the target document can be determined based on the authenticity results corresponding to each static anti-counterfeiting point and the authenticity results corresponding to each dynamic anti-counterfeiting point.

[0045] In one embodiment of the present application, after the server 103 identifies the authenticity of the target certificate 104 , it can return the identification result to the terminal 101 through the network 102 , and then the terminal 101 can display the identification result of the target certificate 104 to the user.

[0046] The document authenticity identification method provided in the embodiment of the present application is generally executed by server 103. Server 103 is used to receive multiple document images of the target document 104 uploaded by terminal 101, and to identify the authenticity of the target document 104 based on the multiple document images. Accordingly, the document authenticity identification device is generally provided in server 103. However, it will be readily understood by those skilled in the art that the document authenticity identification method provided in the embodiment of the present application can also be executed by terminal 101, and accordingly, the document authenticity identification device can also be provided in terminal 101, and this is not particularly limited in this exemplary embodiment. For example, in one exemplary embodiment, terminal 101 is used to collect and obtain multiple document images of the target document 104, and to identify the authenticity of the target document 104 based on the document authenticity identification method proposed in this application.

[0047] It should be noted that if the document authenticity identification method provided in the embodiment of the present application is executed by the server 103, then the client running on the terminal 101 can implement document authentication by calling a request service via a web page based on a user trigger; if the document authenticity identification method provided in the embodiment of the present application is executed by the terminal 101, then the client running on the terminal 101 can provide the document authentication function by calling a authentication SDK (Software Development Kit). Of course, the client can also call the SDK to determine whether the captured document image is standardized, and combine it with the authentication service provided by the server 103 to quickly identify the authenticity of the document.

[0048] The following is a detailed description of the implementation details of the technical solution of the embodiment of the present application:

[0049] Figure 2 A flowchart of a method for identifying a certificate authenticity according to an embodiment of the present application is shown. The method for identifying a certificate authenticity can be performed by a device having a computing and processing function, such as Figure 1 Refer to the server 103 or terminal 101 shown in Figure 2 As shown, the document authenticity identification method includes at least steps S210 to S240, which are described in detail as follows:

[0050] In step S210, dynamic anti-counterfeiting dots and static anti-counterfeiting dots included in a plurality of document images of a target document are detected. The plurality of document images are obtained by capturing images of the target document at different angles.

[0051] In one embodiment of the present application, the target document may be a physical document such as an ID card, passport, social security card, or medical insurance card. The dynamic security point may be an anti-counterfeiting mark whose attribute information changes when the document is in different postures. For example, color-changing ink, dynamic printing, hologram, laser imaging, three-dimensional relief, etc. When the document is in different postures, the dynamic security point on the document will undergo certain changes, and the changes in the dynamic security point follow a certain regular change. For example, the color, clarity, texture, or shape of the dynamic security point on the document will change. Static security points are anti-counterfeiting marks that do not change due to different angles of the document or different lighting conditions.

[0052] Generally speaking, forged documents usually do not contain anti-counterfeiting points, or only contain some anti-counterfeiting points, or the anti-counterfeiting points are different from those of real documents (for example, the dynamic anti-counterfeiting points on the forged document will not change due to changes in angle and lighting, or the way they change is different from that of real documents, etc.). Therefore, multiple document images can be obtained by capturing the target document at different angles to comprehensively determine the authenticity of the document based on the dynamic anti-counterfeiting points and static anti-counterfeiting points.

[0053] In one embodiment of the present application, an image capture device may capture a video stream containing a target document from different angles and then obtain multiple document images from the video stream; or it may directly capture multiple document images of the target document from different capture angles. Specifically, upon detecting a document image capture request, the image capture device may activate a camera to capture images of the target document from different angles, thereby obtaining video streams or multiple document images captured from different angles.

[0054] In one embodiment of the present application, Figure 3 As shown, when capturing an image of a target document, the target document may be flipped at different angles and flipped to different positions (such as Figure 3 Image capture is performed on the target document at positions A, B, C, and D shown in the figure, thereby obtaining a video stream or multiple document images.

[0055] In one embodiment of the present application, Figure 4 As shown, the method of capturing images of the target document at different angles can also be to place the target document at a fixed position and move the image capture device 401 to different positions (such as Figure 3 At positions a, b, c, and d shown in FIG, images of the target document are captured at different capture angles, thereby obtaining a video stream or multiple document images.

[0056] In addition, in other embodiments of the present application, the positions of the target document and the camera device can be changed to capture a video stream or multiple document images.

[0057] In one embodiment of the present application, when capturing a video stream or multiple ID images of a target ID, the position of the target ID contained in the video image frame in the captured video stream or the captured ID image can also be detected. If the detected position of the target ID does not meet the set conditions, the capture can be repeated until the position of the target ID contained in the captured video image frame or ID image meets the set conditions. Alternatively, if the detected position of the target ID is outside the video image frame or ID image, or if the detected target ID is too small or too large in the video image frame or ID image (the small or large size is caused by the distance between the camera and the target ID being too far or too close during shooting), the user can be prompted to recapture the image.

[0058] In one embodiment of the present application, the position of the target document contained in the video image frame can be detected in the following manner (since the method for detecting the position of the target document in the document image is similar, the following description takes the detection of the position of the target document contained in the video image frame as an example):

[0059] Downsampling the video image frame by sequentially connecting a plurality of convolution blocks, wherein a first convolution block in the plurality of convolution blocks is used to downsample the video image frame, and an i+1th convolution block in the plurality of convolution blocks is used to downsample the output feature map of the ith convolution block, where i is greater than 0;

[0060] Sequentially upsample the corresponding feature maps of the i+1th convolution block in the multiple convolution blocks, merge the upsampling result with the output feature map of the i-th convolution block, and use the merged result as the corresponding feature map of the i-th convolution block, wherein the upsampling process and the downsampling process have the same sampling scale;

[0061] The position of the target document contained in the video image frame is identified according to the corresponding feature map of the first convolution block in the plurality of convolution blocks. It should be noted that the corresponding feature map of the last convolution block in the plurality of convolution blocks is the output feature map of the last convolution block.

[0062] Specifically, if Figure 5As shown, for a video image frame 500, a convolution block is used to perform convolution downsampling by 2×2 (the specific scale is only an example) times to obtain a feature map 501 (feature map 501 is the output feature map of the first convolution block), and then the convolution downsampling by 2×2 is continued to obtain a feature map 502 (feature map 502 is the output feature map of the second convolution block), and so on to obtain feature maps 503 and feature maps 504 (feature map 503 is the output feature map of the third convolution block, and feature map 504 is the output feature map of the fourth convolution block. This embodiment is illustrated by taking four convolution blocks as an example). The corresponding feature map 504' of the 4th convolution block (that is, the last convolution block in the 4 sequentially connected convolution blocks) is upsampled by 2×2 (since the 4th convolution block is the last convolution block, the corresponding feature map of the 4th convolution block is the output feature map of the 4th convolution block) and merged with the output feature map 503 of the 3rd convolution block to obtain the corresponding feature map 503' of the 3rd convolution block. Then, the corresponding feature map 503' of the 3rd convolution block is upsampled by 2×2 and merged with the output feature map 503 of the 2nd convolution block. The output feature map 502 of the second convolution block is merged to obtain the corresponding feature map 502' of the second convolution block. The corresponding feature map 502' of the second convolution block is upsampled by 2×2 and merged with the output feature map 501 of the first convolution block to obtain the corresponding feature map 501' of the first convolution block. The feature map 501' is a feature map of the same size as the video image frame 500. Finally, the feature map is input into the sigmoid function to obtain the edge segmentation result of the target document, that is, the position of the white frame line in 505 is the position of the detected target document.

[0063] In one embodiment of the present application, when capturing a video stream or multiple ID images of a target ID, the flip angle of the target ID contained in the captured video image frames or ID images can also be detected. If the detected flip angle of the target ID does not meet a set condition, the capture can be repeated until the flip angle of the target ID contained in the captured video image frames or ID images meets the set condition. Optionally, if the detected flip angle of the target ID is too large or too small, the user can be prompted to recapture.

[0064] In one embodiment of the present application, the flip angle of the target document contained in the video image frame in the video stream is detected (since the method of detecting the flip angle of the target document in the document picture is similar, the following description is made by taking the detection of the flip angle of the target document contained in the video image frame as an example). The target document detection can be performed in the video image frame contained in the video stream to identify the document detection frame containing the target document in the video image frame, and then the flip angle of the target document contained in the video image frame is determined based on the side length ratio of the document detection frame identified in the video image frame.

[0065] For example, the ratio of the left length of the ID detection frame to the left length of the target ID can be calculated and the inverse cosine function can be used to obtain the first angle. Similarly, the ratio of the right length of the ID detection frame to the right length of the target ID can be calculated and the inverse cosine function can be used to obtain the second angle. The sum of the first angle and the second angle is averaged to obtain the flipping angle of the target ID.

[0066] In one embodiment, the ratio of the length of two adjacent sides of the document detection frame and the ratio of the length of two adjacent sides of the target document can also be calculated, and then the flip angle of the target document contained in each video image frame can be determined based on the corresponding relationship between the ratio of the length of two adjacent sides of the document detection frame and the ratio of the length of two adjacent sides of the target document. For example, the ratio relationship between the ratio of the length of two adjacent sides of the document detection frame and the ratio of the length of two adjacent sides of the target document is calculated, and the flip angle corresponding to the ratio relationship is determined based on the ratio relationship. The method of determining the flip angle based on the corresponding relationship between the ratio of the length of two adjacent sides of the target document and the ratio of the length of two adjacent sides of the document detection frame in this embodiment is more robust and will not be affected by the different sizes of the target documents in different video image frames.

[0067] In one embodiment of the present application, an image detection model can be used to detect each video image frame. The training samples of the image detection model include video image frame samples annotated with document detection frames and enhanced images obtained by performing image enhancement processing on the video image frame samples. Alternatively, the video image frames contained in the video stream can be input into the image detection model, and the image detection model can be used to detect each video image frame to obtain a document detection frame containing the target document in each video image frame.

[0068] In one embodiment of the present application, the image detection model can use PVAnet (Performance VsAccuracy Net), which can maintain detection accuracy while having a faster speed. During the training process for PVAnet, the selected training samples include: video image frame samples marked with document detection frames, and enhanced images obtained by performing image enhancement processing on the video image frame samples. Optionally, the enhanced image can be an image obtained by performing image enhancement processing, such as performing enhancement processing on the video image frame samples using methods such as rotation, brightness, contrast, and noise addition.

[0069] Continue to refer to Figure 2 As shown, in step S220, the image feature information of the location of the static anti-counterfeiting point is extracted to obtain the static anti-counterfeiting point feature, and the image feature information of the location of the dynamic anti-counterfeiting point and the change feature information of the dynamic anti-counterfeiting point between multiple document images are extracted to obtain the dynamic anti-counterfeiting point feature.

[0070] In one embodiment of the present application, the feature information of the static anti-counterfeiting point is the two-dimensional image feature information. Therefore, the image feature information of the location of the static anti-counterfeiting point can be extracted as the static anti-counterfeiting point feature. For example, the static anti-counterfeiting point feature can be extracted by the SIFT (Scale-invariant feature transform) algorithm, or the HOG (Histogram of Oriented Gradient) feature can be extracted as the static anti-counterfeiting point feature, etc., or the static anti-counterfeiting point feature can be extracted by a two-dimensional convolutional neural network. For dynamic anti-counterfeiting points, it is necessary not only to consider the image feature information of the location of the static anti-counterfeiting point, but also to add a time series dimension on this basis to capture the changing characteristics of the dynamic anti-counterfeiting point between multiple frames of the video. Specifically, a three-dimensional convolutional network can be used to extract the dynamic anti-counterfeiting point feature. See below for details.

[0071] In step S230 , the authenticity result corresponding to each static anti-counterfeiting point is identified based on the static anti-counterfeiting point feature, and the authenticity result corresponding to each dynamic anti-counterfeiting point is identified based on the dynamic anti-counterfeiting point feature.

[0072] In one embodiment of the present application, the authenticity result corresponding to each static anti-counterfeiting point can be determined based on the first characteristic value interval of the static anti-counterfeiting point feature of each static anti-counterfeiting point and the authenticity result associated with the first characteristic value interval. For example, the characteristic value interval range of the authenticity and counterfeiting anti-counterfeiting point can be set in advance, and then the authenticity result of the static anti-counterfeiting point can be determined based on the characteristic value interval range of the static anti-counterfeiting point feature.

[0073] Similarly, the authenticity result corresponding to each dynamic anti-counterfeiting point may also be determined according to the second characteristic value interval of the dynamic anti-counterfeiting point feature of each dynamic anti-counterfeiting point and the authenticity result associated with the second characteristic value interval.

[0074] In one embodiment of the present application, authenticity classification methods may also be used to determine the authenticity of static and dynamic security points. For example, an SVM (Support Vector Machine) classifier may be used to classify the authenticity of security points, or a fully connected layer in a neural network may be used to classify the authenticity of security points.

[0075] In one embodiment of the present application, a specified video image frame can be extracted from a video stream containing a target document. The specified video image frame can be all video image frames in the video stream, or can be one or more video image frames extracted from the video stream at set intervals, or can be a video image frame in which the target document is in a horizontal position. Then, image feature information of the location of the static anti-counterfeiting point is extracted from the specified video image frame. Based on this, in one embodiment of the present application, the authenticity of the static anti-counterfeiting point contained in each specified video image frame can be identified based on the static anti-counterfeiting point features extracted from each specified video image frame. Then, based on the authenticity of the static anti-counterfeiting point contained in each specified video image frame, the authenticity of the same static anti-counterfeiting point in each specified video image frame can be determined. Then, based on the authenticity of the same static anti-counterfeiting point in each specified video image frame, the authenticity of each static anti-counterfeiting point on the target document can be calculated.

[0076] Specifically, for any video image frame, the technical solutions described in the aforementioned embodiments can be used to detect the static security points contained therein and identify the authenticity of the static security points. The authenticity results of the same static security points contained in the extracted specified video image frame are then combined to obtain the authenticity results of each static security point. For example, the authenticity result can be an authenticity confidence level. Furthermore, the authenticity confidence levels of the same static security points contained in the specified video image frame can be averaged, and the resulting average value is then used as the authenticity result for each static security point.

[0077] It should be noted that if the authenticity of the static anti-counterfeiting point is determined by directly collecting multiple document images instead of extracting a specified video image frame from the video stream, the specific processing method is similar to the solution based on extracting a specified video image frame from the video stream to determine the authenticity of the static anti-counterfeiting point, and will not be repeated here.

[0078] In one embodiment of the present application, at least one group of video image frames can be extracted from a video stream containing a target document, and then image feature information of the location of the dynamic anti-counterfeiting point and feature information of changes in the dynamic anti-counterfeiting point can be extracted from the at least one group of video image frames. Based on this, in one embodiment of the present application, the authenticity of the dynamic anti-counterfeiting point contained in each group of video image frames can be identified based on the image feature information of the location of the dynamic anti-counterfeiting point and feature information of changes in the dynamic anti-counterfeiting point extracted from each group of video image frames, and then the authenticity of the same dynamic anti-counterfeiting point in each group of video image frames can be determined, and then the authenticity of each dynamic anti-counterfeiting point on the target document can be calculated based on the authenticity of the same dynamic anti-counterfeiting point in each group of video image frames.

[0079] Specifically, for any set of video image frames, the technical solutions described in the aforementioned embodiments can be used to detect the dynamic security points contained therein and identify the authenticity of the dynamic security points. The authenticity results for the same dynamic security points contained in each extracted set of video image frames are then combined to obtain the authenticity results for each dynamic security point. For example, the authenticity result can be an authenticity confidence level. Furthermore, the authenticity confidence levels for the same dynamic security points contained in each set of video image frames can be averaged, and the resulting average value is then used as the authenticity result for each dynamic security point.

[0080] Of course, if at least one group of video image frames is not extracted from the video stream to determine the authenticity of the dynamic anti-counterfeiting point, but multiple document images are directly collected to determine the authenticity of the dynamic anti-counterfeiting point, then the specific processing method is similar to the scheme based on extracting at least one group of video image frames from the video stream to determine the authenticity of the dynamic anti-counterfeiting point. For example, at least one group of document images can be obtained based on these multiple document images, and then authenticity identification can be performed. I will not go into details.

[0081] In one embodiment of the present application, as described in the previous embodiment, a three-dimensional convolutional network can be used to extract dynamic anti-counterfeiting point features, and the authenticity results corresponding to the dynamic anti-counterfeiting points can also be output. Specifically, multiple ID image frames containing dynamic anti-counterfeiting points (these multiple ID images can be extracted from a video stream containing the target ID) can be input into the three-dimensional convolutional network to extract a multi-dimensional feature map of the dynamic anti-counterfeiting points. The multi-dimensional feature map is then converted into a one-dimensional feature map through a downsampling unit in the three-dimensional convolutional network, and the authenticity results corresponding to the dynamic anti-counterfeiting points are output through the fully connected layer in the three-dimensional convolutional network.

[0082] Alternatively, a schematic diagram of the authenticity result corresponding to the dynamic anti-counterfeiting point identified by the three-dimensional convolutional network can be referred to Figure 6 As shown, as an example, four convolution blocks can be used. These four convolution blocks are serialized in sequence. Each convolution block is a structure of conv3d+BN (i.e. Batch Normalization) layer+Relu (Rectified Linear Unit). The feature units of the four convolution blocks correspond to Figure 6 4 columns in . Figure 6 The dotted line connection between the convolution features is used to represent a feature unit of the next layer (i.e. Figure 6The cube in the figure represents a unit in the feature map of this layer, and the feature unit of the three-dimensional convolution is obtained by convolving several adjacent feature units in the previous layer. Finally, the multidimensional feature map is converted into a one-dimensional feature vector 601, and the authenticity confidence level corresponding to the security point is output through a fully connected layer. The authenticity confidence level can be in the form of: authenticity + confidence level, or falsehood + confidence level. For example, authenticity is 99.3%; falsehood is 0.7%.

[0083] Continue to refer to Figure 2 As shown, in step S240, the authenticity of the target document is determined according to the authenticity results corresponding to each static anti-counterfeiting point and the authenticity results corresponding to each dynamic anti-counterfeiting point.

[0084] In one embodiment of the present application, if the authenticity result includes authenticity confidence, then the authenticity confidence corresponding to each static anti-counterfeiting point and the authenticity confidence corresponding to each dynamic anti-counterfeiting point can be weighted according to the weight of each static anti-counterfeiting point and the weight of each dynamic anti-counterfeiting point to obtain a comprehensive authenticity confidence value, and then the authenticity of the target certificate can be determined based on the comprehensive authenticity confidence value.

[0085] For example, assuming that the target certificate has 1 static anti-counterfeiting point and 2 dynamic anti-counterfeiting points, the weight of the static anti-counterfeiting point a is 0.2, the weight of the dynamic anti-counterfeiting point b is 0.4, and the weight of the dynamic anti-counterfeiting point c is 0.4. The authenticity confidence of the static anti-counterfeiting point a is: 0.4 for authenticity; the authenticity confidence of the dynamic anti-counterfeiting point b is: 0.7 for falsehood; the authenticity confidence of the dynamic anti-counterfeiting point c is 0.5 for authenticity. Then the comprehensive confidence value of the target certificate being authentic can be calculated as 0.2×0.4+0.4×(1-0.7)+0.4×0.5=0.4.

[0086] After calculating the target document's comprehensive authenticity confidence value, the comprehensive authenticity confidence value can be compared with a preset threshold. If the comprehensive authenticity confidence value of the target document is greater than the preset threshold, the target document is determined to be authentic. Conversely, if the comprehensive authenticity confidence value of the target document is less than or equal to the preset threshold, the target document is determined to be false. The preset threshold can be set based on actual circumstances.

[0087] In summary, the technical solution of the embodiment of the present application is mainly based on the dynamic anti-counterfeiting points and static anti-counterfeiting points on the document image to jointly identify the authenticity of the document. For the static anti-counterfeiting points, the image feature information of the location of the static anti-counterfeiting points is taken into account. For the dynamic anti-counterfeiting points, not only the image feature information of the location of the dynamic anti-counterfeiting points is taken into account, but also the changing feature information of the dynamic anti-counterfeiting points between multiple document images is taken into account. The authenticity results corresponding to each static anti-counterfeiting point and the authenticity results corresponding to each dynamic anti-counterfeiting point can be determined first, and then the authenticity of the target document can be identified according to the authenticity results corresponding to each static anti-counterfeiting point and the authenticity results corresponding to each dynamic anti-counterfeiting point, thereby improving the accuracy of document authenticity identification.

[0088] like Figure 7 As shown, in one embodiment of the present application, a method for identifying the authenticity of a certificate may include the following steps:

[0089] Step S701: After the user takes a photo of the ID card, the user detects the position of the ID card in the image.

[0090] In one embodiment of the present application, a user can be instructed to hold an ID and perform a specific motion to capture a video. The specific motion can be such that the ID's dynamic security features undergo significant changes. For example, the dynamic security features of the Hong Kong ID card, version 03, include color-changing ink, a dynamic printed "HK" logo, and a dynamic printed portrait. These dynamic security features significantly change when the ID is flipped upside down. Therefore, the user can be instructed to capture the video starting with the ID facing forward relative to the camera's horizontal position, flipping it upward, then returning it to the normal position, and then flipping it downward.

[0091] Step S702 determines whether the position of the ID in the image meets the requirements. If so, dynamic and static security point identification are performed, and the ID verification result is obtained based on the dynamic and static security point identification results. If the ID position in the image does not meet the requirements, a prompt message is returned, prompting the user to retake the photo.

[0092] In one embodiment of the present application, for a video image frame in a captured video, the position of the document is obtained to determine whether it meets the specifications. To obtain the document position, an algorithm such as Hough line detection can be used to detect the four edges of the document and combine them to obtain the position coordinates of the document; a neural network method can also be used, such as through Figure 5Alternatively, a more accurate target detection method can be used. After obtaining the ID's position, the system can determine whether the ID has moved out of the frame, is too far away, or is too close, to confirm whether the ID image meets specifications. If not, a prompt can be returned to guide the user to retake the image. Furthermore, the ID's flip angle can be obtained to determine whether the flip angle is too large or too small. If so, a prompt can be returned to guide the user to retake the image.

[0093] In one embodiment of the present application, dynamic anti-counterfeiting points change when the document is at different angles or under different lighting conditions. For example, the Bauhinia flower on the 18th edition of the Hong Kong Identity Card changes significantly under different lighting conditions, and the dynamically printed "HK" logo on the 03rd edition of the Hong Kong Identity Card presents different "H" or "K" characters at different angles. Static anti-counterfeiting points, on the other hand, are anti-counterfeiting features that do not change due to different angles or lighting conditions. They are mostly textured text, such as the rainbow-printed anti-counterfeiting points on the 03rd edition of the Hong Kong Identity Card, which use microtext technology. Therefore, it is necessary to identify dynamic and static anti-counterfeiting points separately. Before identification, the feature information of the static and dynamic anti-counterfeiting points must first be extracted. The specific extraction process and the identification process of the anti-counterfeiting points can refer to the technical solutions of the aforementioned embodiments and will not be described in detail.

[0094] It should be noted that since static security points are identified on a single frame, they can be identified on every frame of the video. Alternatively, they can be identified on intervals, and the average value can be taken to obtain the identification result for the static security point, which is a more robust approach. Alternatively, specific frames, such as horizontal frames, can be identified, which is more efficient while ensuring a certain level of accuracy. For dynamic security points, since multiple frames are required to identify the changing characteristics of dynamic feature points, the entire video frame can be input, or segmented, or several groups of random video frames can be randomly selected for identification, and the average value can be taken as the identification result for the dynamic security point.

[0095] After obtaining the identification results for each static and dynamic security point, the corresponding weights can be adjusted based on their importance and effectiveness to calculate the final document authenticity score. For example, all security points on the document can be assigned the same weight, averaging the weights to obtain the final authenticity confidence score. A corresponding threshold can then be set and compared with the final authenticity confidence score to obtain the final authentication result: determining whether the document is authentic or forged.

[0096] The following describes an embodiment of the device of the present application, which can be used to perform the method for identifying the authenticity of a certificate in the above embodiment of the present application. For details not disclosed in the embodiment of the device of the present application, please refer to the embodiment of the method for identifying the authenticity of a certificate in the above embodiment of the present application.

[0097] Figure 8 A block diagram of a device for identifying a certificate authenticity according to an embodiment of the present application is shown. The device for identifying a certificate authenticity can be set in a device with a computing and processing function, such as a Figure 1 In the server 103 or terminal 101 shown in .

[0098] Reference Figure 8 As shown, a device 800 for identifying authenticity of a certificate according to an embodiment of the present application includes: an anti-counterfeiting point detection unit 802 , an extraction unit 804 , a processing unit 806 and a determination unit 808 .

[0099] Among them, the anti-counterfeiting point detection unit 802 is configured to detect dynamic anti-counterfeiting points and static anti-counterfeiting points contained in multiple document images of the target document, and the multiple document images are obtained by capturing images of the target document at different angles; the extraction unit 804 is configured to extract image feature information of the position of the static anti-counterfeiting point to obtain static anti-counterfeiting point features, extract image feature information of the position of the dynamic anti-counterfeiting point and change feature information of the dynamic anti-counterfeiting point between multiple document images to obtain dynamic anti-counterfeiting point features; the processing unit 806 is configured to identify the authenticity results corresponding to each static anti-counterfeiting point based on the static anti-counterfeiting point features, and identify the authenticity results corresponding to each dynamic anti-counterfeiting point based on the dynamic anti-counterfeiting point features; the determination unit 808 is configured to determine the authenticity of the target document based on the authenticity results corresponding to each static anti-counterfeiting point and the authenticity results corresponding to each dynamic anti-counterfeiting point.

[0100] In some embodiments of the present application, based on the aforementioned scheme, the document authenticity identification device 800 further includes: an acquisition unit, a position detection unit and an acquisition unit; wherein, the acquisition unit is configured to acquire a video stream containing the target document from different angles; the position detection unit is configured to detect the position of the target document contained in the video image frame in the video stream; the acquisition unit is further configured to: if the position detection unit detects that the position of the target document does not meet the set conditions, then re-acquire the video stream containing the target document until the position of the target document contained in the acquired video image frame meets the set conditions; the acquisition unit is configured to: if it is detected that the position of the target document meets the set conditions, then acquire the multiple document images from the video image frames contained in the video stream.

[0101] In some embodiments of the present application, based on the aforementioned scheme, the position detection unit is configured to: downsample the video image frame through a plurality of sequentially connected convolution blocks, wherein the first convolution block among the plurality of convolution blocks is used to downsample the video image frame, and the i+1th convolution block among the plurality of convolution blocks is used to downsample the output feature map of the i-th convolution block, where i is greater than 0; upsample the corresponding feature map of the i+1th convolution block among the plurality of convolution blocks in sequence, and merge the result of the upsampling process with the output feature map of the i-th convolution block, and use the merged result as the corresponding feature map of the i-th convolution block, the upsampling process and the downsampling process have the same sampling scale; and identify the position of the target document contained in the video image frame according to the corresponding feature map of the first convolution block among the plurality of convolution blocks.

[0102] In some embodiments of the present application, based on the aforementioned scheme, the document authenticity identification device 800 further includes: an angle detection unit, configured to detect the flip angle of the target document contained in the video image frame in the video stream; the acquisition unit is further configured to: if the detected flip angle of the target document does not meet the set conditions, re-acquire the video stream containing the target document until the flip angle of the target document contained in the acquired video image frame meets the set conditions.

[0103] In some embodiments of the present application, based on the aforementioned scheme, the angle detection unit is configured to: perform target document detection in the video image frame contained in the video stream to identify the document detection frame containing the target document in the video image frame; and determine the flip angle of the target document contained in the video image frame based on the side length ratio of the document detection frame identified in the video image frame.

[0104] In some embodiments of the present application, based on the aforementioned scheme, the extraction unit 804 extracts the image feature information of the location of the dynamic anti-counterfeiting point and the change feature information of the dynamic anti-counterfeiting point between multiple document images to obtain the dynamic anti-counterfeiting point feature, and the processing unit 806 identifies the authenticity results corresponding to each dynamic anti-counterfeiting point based on the dynamic anti-counterfeiting point feature, including: inputting multiple document image frames containing the dynamic anti-counterfeiting point into a three-dimensional convolutional network to extract a multidimensional feature map of the dynamic anti-counterfeiting point through the three-dimensional convolutional network; converting the multidimensional feature map into a one-dimensional feature map through a downsampling unit in the three-dimensional convolutional network, and outputting the authenticity results corresponding to the dynamic anti-counterfeiting point through a fully connected layer in the three-dimensional convolutional network.

[0105] In some embodiments of the present application, based on the aforementioned scheme, the processing unit 806 is configured to: determine the authenticity result corresponding to each static anti-counterfeiting point according to the first characteristic value interval of the static anti-counterfeiting point feature of each static anti-counterfeiting point, and the authenticity result associated with the first characteristic value interval; determine the authenticity result corresponding to each dynamic anti-counterfeiting point according to the second characteristic value interval of the dynamic anti-counterfeiting point feature of each dynamic anti-counterfeiting point, and the authenticity result associated with the second characteristic value interval.

[0106] In some embodiments of the present application, based on the aforementioned scheme, the multiple document images include: specified video image frames extracted from a video stream containing the target document, and the specified video image frames include any of the following: each video image frame in the video stream, one or more video image frames extracted from the video stream at a set interval, and a video image frame in which the target document is in a horizontal position; the extraction unit 804 is configured to extract image feature information of the location of the static anti-counterfeiting point in the specified video image frame.

[0107] In some embodiments of the present application, based on the aforementioned scheme, the processing unit 806 is configured to: identify the authenticity of the static anti-counterfeiting points contained in each designated video image frame based on the static anti-counterfeiting point features extracted from each designated video image frame; determine the authenticity of the same static anti-counterfeiting point in each designated video image frame based on the authenticity of the static anti-counterfeiting points contained in each designated video image frame; calculate the authenticity of each static anti-counterfeiting point on the target document based on the authenticity of the same static anti-counterfeiting point in each designated video image frame.

[0108] In some embodiments of the present application, based on the aforementioned scheme, the multiple document images include: at least one group of video image frames extracted from a video stream containing the target document; the extraction unit 804 is configured to: extract image feature information of the location of the dynamic anti-counterfeiting point and change feature information of the dynamic anti-counterfeiting point from the at least one group of video image frames.

[0109] In some embodiments of the present application, based on the aforementioned scheme, the processing unit 806 is configured as follows: if multiple groups of video image frames are extracted from the video stream containing the target certificate, then based on the image feature information of the position of the dynamic anti-counterfeiting point extracted from each group of video image frames and the change feature information of the dynamic anti-counterfeiting point, the authenticity of the dynamic anti-counterfeiting point contained in each group of video image frames is identified; based on the authenticity of the dynamic anti-counterfeiting point contained in each group of video image frames, the authenticity of the same dynamic anti-counterfeiting point in each group of video image frames is determined; based on the authenticity of the same dynamic anti-counterfeiting point in each group of video image frames, the authenticity of each dynamic anti-counterfeiting point on the target certificate is calculated.

[0110] In some embodiments of the present application, based on the aforementioned scheme, the authenticity result includes authenticity confidence; the determination unit 808 is configured to: weight the authenticity confidence corresponding to each static anti-counterfeiting point and the authenticity confidence corresponding to each dynamic anti-counterfeiting point according to the weight of each static anti-counterfeiting point and the weight of each dynamic anti-counterfeiting point to obtain a comprehensive authenticity confidence value; determine the authenticity of the target certificate according to the comprehensive authenticity confidence value.

[0111] Figure 9 A schematic diagram of the structure of a computer system suitable for implementing an electronic device according to an embodiment of the present application is shown.

[0112] It should be noted that Figure 9 The computer system 900 of the electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0113] like Figure 9 As shown, the computer system 900 includes a central processing unit (CPU) 901, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 902 or the program loaded from the storage part 908 into the random access memory (RAM) 903, such as the method described in the above embodiment. Various programs and data required for system operation are also stored in the RAM 903. The CPU 901, ROM 902 and RAM 903 are connected to each other via a bus 904. An input / output (I / O) interface 905 is also connected to the bus 904.

[0114] The following components are connected to the I / O interface 905: an input section 906 including a keyboard, a mouse, and the like; an output section 907 including devices such as a cathode ray tube (CRT), a liquid crystal display (LCD), and a speaker; a storage section 908 including a hard disk and the like; and a communication section 909 including a network interface card such as a LAN (Local Area Network) card or a modem. The communication section 909 performs communication processing via a network such as the Internet. A drive 910 is also connected to the I / O interface 905 as needed. Removable media 911, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, and the like, are installed in the drive 910 as needed, so that computer programs read therefrom can be installed into the storage section 908 as needed.

[0115] In particular, according to an embodiment of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 909, and / or installed from a removable medium 911. When the computer program is executed by the central processing unit (CPU) 901, the various functions defined in the system of the present application are executed.

[0116] It should be noted that the computer-readable medium shown in the embodiments of the present application may be a computer-readable signal medium or a computer-readable storage medium or any combination of the two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or component, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present application, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries a computer-readable computer program. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. A computer program embodied on a computer-readable medium may be transmitted using any suitable medium, including but not limited to wireless, wired, or any suitable combination thereof.

[0117] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present application. Among them, each box in the flowchart or block diagram can represent a module, program segment, or part of the code, and the above-mentioned module, program segment, or part of the code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0118] The units involved in the embodiments described in this application may be implemented by software or hardware, and the units described may also be set in a processor. In some cases, the names of these units do not constitute limitations on the units themselves.

[0119] As another aspect, the present application further provides a computer-readable medium, which may be included in the electronic device described in the above embodiments, or may exist independently without being incorporated into the electronic device. The computer-readable medium carries one or more programs, and when the one or more programs are executed by the electronic device, the electronic device implements the method described in the above embodiments.

[0120] It should be noted that, although several modules or units of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiment of the application, the features and functions of two or more modules or units described above can be concretized in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into multiple modules or units to be concretized.

[0121] Through the description of the above embodiments, it is easy for those skilled in the art to understand that the example embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solution according to the embodiments of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which can be a personal computer, a server, a touch terminal, or a network device, etc.) to execute the method according to the embodiments of the present application.

[0122] Those skilled in the art will readily conceive of other embodiments of the present application after considering the specification and practicing the embodiments disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present application that follow the general principles of this application and include common knowledge or customary techniques in the art that are not disclosed herein.

[0123] It should be understood that the present application is not limited to the exact structures described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A method for identifying the authenticity of a certificate, characterized in that: include: Capture video streams containing target documents from different angles; Detecting a position of a target document contained in a video image frame in the video stream; If it is detected that the position of the target document does not meet the set conditions, re-capturing the video stream containing the target document until the position of the target document contained in the captured video image frame meets the set conditions; Performing target document detection in the video image frames included in the video stream to identify a document detection frame including the target document in the video image frames; determining a flip angle of a target document contained in the video image frame according to a side length ratio of a document detection frame identified in the video image frame; If the detected flip angle of the target document does not meet the set condition, re-capturing the video stream containing the target document until the flip angle of the target document contained in the captured video image frame meets the set condition; If it is detected that the position and flip angle of the target document meet corresponding set conditions, a plurality of document images are obtained from the video image frames included in the video stream; Detecting dynamic anti-counterfeiting points and static anti-counterfeiting points contained in a plurality of document images of a target document, wherein the plurality of document images are obtained by capturing images of the target document at different angles; Extracting image feature information of the location of the static anti-counterfeiting point to obtain a static anti-counterfeiting point feature; extracting image feature information of the location of the dynamic anti-counterfeiting point and feature information of changes in the dynamic anti-counterfeiting point between multiple document images to obtain a dynamic anti-counterfeiting point feature; Identify the authenticity of each static anti-counterfeiting point based on the static anti-counterfeiting point features, and identify the authenticity of each dynamic anti-counterfeiting point based on the dynamic anti-counterfeiting point features; The authenticity of the target certificate is determined according to the authenticity results corresponding to the static anti-counterfeiting points and the authenticity results corresponding to the dynamic anti-counterfeiting points.

2. The method for identifying the authenticity of a certificate according to claim 1, wherein: Detecting a position of a target certificate contained in a video image frame in the video stream includes: Downsampling the video image frame by sequentially connecting a plurality of convolution blocks, wherein a first convolution block among the plurality of convolution blocks is used to downsample the video image frame, and an (i+1)th convolution block among the plurality of convolution blocks is used to downsample an output feature map of an (i)th convolution block, where (i) is greater than 0; Sequentially upsampling the corresponding feature maps of the i+1th convolution block in the multiple convolution blocks, merging the upsampling result with the output feature map of the i-th convolution block, and using the merged result as the corresponding feature map of the i-th convolution block, wherein the upsampling process has the same sampling scale as the downsampling process; The position of the target document contained in the video image frame is identified according to the corresponding feature map of the first convolution block among the multiple convolution blocks.

3. The method for identifying the authenticity of a certificate according to claim 1, wherein: Extracting image feature information of the location of the dynamic anti-counterfeiting point and feature information of changes in the dynamic anti-counterfeiting point between multiple document images to obtain dynamic anti-counterfeiting point features, and identifying authenticity results corresponding to each dynamic anti-counterfeiting point based on the dynamic anti-counterfeiting point features, including: Inputting a plurality of document images containing the dynamic anti-counterfeiting dots into a three-dimensional convolutional network to extract a multi-dimensional feature map of the dynamic anti-counterfeiting dots through the three-dimensional convolutional network; The multidimensional feature map is converted into a one-dimensional feature map through a downsampling unit in the three-dimensional convolutional network, and the authenticity result corresponding to the dynamic anti-counterfeiting point is output through a fully connected layer in the three-dimensional convolutional network.

4. The method for identifying the authenticity of a certificate according to claim 1, wherein: Identifying the authenticity results corresponding to the respective static anti-counterfeiting points based on the static anti-counterfeiting point features, including: determining the authenticity results corresponding to the respective static anti-counterfeiting points based on a first characteristic value interval in which the static anti-counterfeiting point features of the respective static anti-counterfeiting points are located, and the authenticity results associated with the first characteristic value interval; Identifying the authenticity results corresponding to each dynamic anti-counterfeiting point based on the dynamic anti-counterfeiting point features includes: determining the authenticity results corresponding to each dynamic anti-counterfeiting point based on the second characteristic value interval in which the dynamic anti-counterfeiting point features of each dynamic anti-counterfeiting point are located, and the authenticity results associated with the second characteristic value interval.

5. The method for identifying the authenticity of a certificate according to claim 1, wherein: The multiple document images include: specified video image frames extracted from a video stream containing the target document, the specified video image frames including any one of the following: each video image frame in the video stream, one or more video image frames extracted from the video stream at a set interval, and a video image frame in which the target document is in a horizontal position; Extracting the image feature information of the location of the static anti-counterfeiting point includes: extracting the image feature information of the location of the static anti-counterfeiting point in the designated video image frame.

6. The method for identifying the authenticity of a certificate according to claim 5, characterized in that: Identifying the authenticity of each static anti-counterfeiting point based on the static anti-counterfeiting point features includes: Based on the static anti-counterfeiting point features extracted from each designated video image frame, identifying the authenticity of the static anti-counterfeiting point contained in each designated video image frame; Determining the authenticity of the same static anti-counterfeiting point in each designated video image frame based on the authenticity of the static anti-counterfeiting point contained in each designated video image frame; The authenticity results of each static anti-counterfeiting point on the target document are calculated according to the authenticity results of the same static anti-counterfeiting point in each designated video image frame.

7. The method for identifying the authenticity of a certificate according to claim 1, wherein: The plurality of document images include: at least one set of video image frames extracted from a video stream containing the target document; Extracting image feature information of the location of the dynamic anti-counterfeiting point and feature information of changes in the dynamic anti-counterfeiting point between multiple document images, including: Image feature information of the location of the dynamic anti-counterfeiting point and change feature information of the dynamic anti-counterfeiting point are extracted from the at least one group of video image frames.

8. The method for identifying the authenticity of a certificate according to claim 7, characterized in that: If multiple groups of video image frames are extracted from a video stream containing the target certificate, the authenticity results corresponding to the static anti-counterfeiting points are identified based on the static anti-counterfeiting point features, including: Based on the image feature information of the positions of the dynamic anti-counterfeiting dots extracted from each set of video image frames and the change feature information of the dynamic anti-counterfeiting dots, the authenticity of the dynamic anti-counterfeiting dots contained in each set of video image frames is identified; Determining the authenticity of the same dynamic anti-counterfeiting point in each group of video image frames according to the authenticity of the dynamic anti-counterfeiting point contained in each group of video image frames; The authenticity results of each dynamic anti-counterfeiting point on the target document are calculated according to the authenticity results of the same dynamic anti-counterfeiting point in each group of video image frames.

9. The method for identifying the authenticity of a certificate according to any one of claims 1 to 8, characterized in that: The authenticity result includes authenticity confidence; Determining the authenticity of the target document according to the authenticity results corresponding to the static anti-counterfeiting points and the authenticity results corresponding to the dynamic anti-counterfeiting points includes: According to the weights of the static anti-counterfeiting points and the dynamic anti-counterfeiting points, weighting the authenticity confidences corresponding to the static anti-counterfeiting points and the authenticity confidences corresponding to the dynamic anti-counterfeiting points to obtain a comprehensive authenticity confidence value; The authenticity of the target certificate is determined based on the comprehensive authenticity confidence value.

10. A device for identifying the authenticity of a certificate, characterized in that: include: a collection unit configured to collect video streams containing a target document from different angles; a position detection unit configured to detect a position of a target document contained in a video image frame in the video stream; The acquisition unit is further configured to recapture the video stream containing the target document if the position detection unit detects that the position of the target document does not meet the set condition, until the position of the target document included in the captured video image frame meets the set condition; an angle detection unit configured to detect a target document in a video image frame included in the video stream, so as to identify a document detection frame including the target document in the video image frame; and determine a flip angle of the target document included in the video image frame based on a side length ratio of the document detection frame identified in the video image frame; The acquisition unit is further configured to, if the detected flip angle of the target document does not meet the set condition, re-acquire the video stream containing the target document until the flip angle of the target document contained in the acquired video image frame meets the set condition; an acquisition unit configured to acquire a plurality of document images from the video image frames included in the video stream if it is detected that the position and flip angle of the target document meet corresponding set conditions; a detection unit configured to detect dynamic anti-counterfeiting points and static anti-counterfeiting points included in the plurality of document images of the target document, the plurality of document images being obtained by capturing images of the target document at different angles; an extraction unit configured to extract image feature information of the location of the static anti-counterfeiting dot to obtain a static anti-counterfeiting dot feature, and extract image feature information of the location of the dynamic anti-counterfeiting dot and feature information of changes in the dynamic anti-counterfeiting dot between multiple document images to obtain a dynamic anti-counterfeiting dot feature; A processing unit configured to identify the authenticity of each static anti-counterfeiting point based on the static anti-counterfeiting point features, and to identify the authenticity of each dynamic anti-counterfeiting point based on the dynamic anti-counterfeiting point features; The determining unit is configured to determine the authenticity of the target document according to the authenticity results corresponding to the static anti-counterfeiting points and the authenticity results corresponding to the dynamic anti-counterfeiting points.

11. A computer-readable medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the document authenticity identification method according to any one of claims 1 to 9 is implemented.

12. An electronic device, characterized in that: include: one or more processors; A storage device for storing one or more programs, which, when executed by the one or more processors, enables the one or more processors to implement the document authenticity identification method as described in any one of claims 1 to 9.

13. A computer program product, characterized in that The computer program product includes computer instructions, which are stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the document authenticity identification method as described in any one of claims 1 to 9.

Images