CN110570209A

CN110570209A - Certificate authenticity verification method and device, computer equipment and storage medium

Info

Publication number: CN110570209A
Application number: CN201910696321.2A
Authority: CN
Inventors: 苏智辉; 郭玲玲
Original assignee: Ping An Technology Shenzhen Co Ltd
Current assignee: Ping An Technology Shenzhen Co Ltd
Priority date: 2019-07-30
Filing date: 2019-07-30
Publication date: 2019-12-13
Also published as: WO2021017610A1

Abstract

The application discloses a certificate authenticity verification method, a certificate authenticity verification device, computer equipment and a storage medium, wherein the certificate authenticity verification method comprises the following steps: acquiring image information of a target certificate at different continuous shooting angles, and performing characteristic detection on the image information at each shooting angle to acquire a plurality of continuous characteristic information, wherein the characteristic detection comprises detecting color-changing ink values of optical color-changing ink areas in the image information, and the characteristic information comprises the color-changing ink values; calculating the difference value between every two adjacent color-changing ink values and taking an absolute value to obtain a plurality of absolute values of the color-changing ink; judging whether a first color-changing ink absolute value larger than a preset color-changing ink absolute value exists in the color-changing ink absolute values or not; if yes, outputting the information that the target certificate is true. The certificate authenticity verification method can solve the problems of low efficiency and low accuracy in the prior art of manually verifying authenticity of the identity card.

Description

certificate authenticity verification method and device, computer equipment and storage medium

Technical Field

the present application relates to the field of computer technologies, and in particular, to a method and an apparatus for verifying authenticity of a certificate, a computer device, and a storage medium.

background

the identity card is used as a legal certificate for proving the personal identity of citizens, plays an important role in application scenes such as banks, customs, airports, railway stations and the like, and more occasions provide higher requirements for the high efficiency and accuracy of identity card verification.

in the prior art, in order to verify the authenticity of the identity card, generally, related workers are required to perform consistency comparison on information read from a chip and information derived from an identity information database, or similarity comparison is performed by using a face photo read from the chip and a photo printed on the surface of the identity card, so as to determine the authenticity of the identity card.

disclosure of Invention

The application mainly aims to provide a certificate authenticity verification method, a certificate authenticity verification device, computer equipment and a storage medium, and aims to solve the problems of low efficiency and low accuracy in the prior art of manually verifying authenticity of an identity card.

The application provides a certificate authenticity verification method, which comprises the following steps:

Acquiring image information of a target certificate at different continuous shooting angles, and performing characteristic detection on the image information at each shooting angle to acquire a plurality of continuous characteristic information, wherein the characteristic detection comprises detecting color-changing ink values of optical color-changing ink areas in the image information, and the characteristic information comprises the color-changing ink values;

Calculating the difference value between every two adjacent color-changing ink values and taking an absolute value to obtain a plurality of absolute values of the color-changing ink;

judging whether a first color-changing ink absolute value larger than a preset color-changing ink absolute value exists in the color-changing ink absolute values or not;

And if the first color-changing ink absolute value which is larger than the preset color-changing ink absolute value exists in the plurality of color-changing ink absolute values, outputting the information that the target certificate is true.

further, the characteristic detection further includes the character detection by using a preset character recognition neural network model, the characteristic information further includes a character detection result, the image information of the target certificate under different continuous shooting angles is obtained, the characteristic detection is performed on the image information under each shooting angle, and after the step of obtaining a plurality of continuous characteristic information, the method further includes:

Determining the certificate type of the target certificate according to the detection results of the plurality of characters;

and according to the certificate type, determining a preset color-changing ink absolute value corresponding to the certificate type from the preset certificate type-color-changing ink absolute value corresponding relation.

Further, the characteristic detection further includes character detection of an optical character area in the image information by using a preset character recognition neural network model, the characteristic information further includes a character detection result, and before the step of outputting the information that the target certificate is true, the method further includes:

judging whether a first character detection result meeting a preset character condition exists in a plurality of character detection results;

and if a first character detection result meeting the preset character condition exists in the plurality of character detection results, executing the step of outputting the information that the target certificate is true.

further, the characteristic detection further includes performing color detection on an optical color-changing ink area in the image information by using a preset color recognition neural network model, the characteristic information further includes a color detection result, and before the step of outputting the information that the target certificate is true, the method further includes:

judging whether a first color detection result meeting a preset color condition exists in the multiple color detection results;

and if a first color detection result meeting the preset color condition exists in the plurality of character detection results, executing a step of outputting information that the target certificate is true.

further, before the steps of obtaining image information of the target certificate at different continuous shooting angles, and performing feature detection on the image information at each shooting angle to obtain a plurality of continuous feature information, the method further comprises:

acquiring a plurality of true certificate images with different inclination angles and carrying character marking information;

Inputting a plurality of true certificate images into a character recognition neural network model for character detection, and outputting a plurality of character information;

Comparing each character information with corresponding character marking information one by one to verify whether the detection precision of the character recognition neural network model reaches the preset precision or not;

and if the detection precision of the character recognition neural network model does not reach the preset precision, repeatedly iterating parameters in the character recognition neural network model until the detection precision of the character recognition neural network model reaches the preset precision.

further, when the image information is a video stream, acquiring the image information of the target certificate at different continuous shooting angles, and performing feature detection on the image information at each shooting angle to acquire a plurality of continuous feature information, including:

Continuously shooting the front side of the target certificate from different shooting angles to obtain a video stream;

and carrying out feature detection on the video stream frame by frame or interval preset frame number to obtain a plurality of continuous feature information.

Further, the step of continuously shooting the front side of the target certificate from different shooting angles to obtain a video stream includes:

Focusing the target certificate, and acquiring a first shape parameter of the focused target certificate, wherein the first shape parameter comprises first length parameters of two opposite edges of the target certificate;

Calculating a deviation value between the two first length parameters, and judging whether the deviation value is smaller than a first preset threshold value or not;

If the deviation value is smaller than a first preset threshold value, continuously shooting the front side of the target certificate from different shooting angles, and acquiring second shape parameters of the target certificate in real time in the shooting process, wherein the second shape parameters comprise second length parameters of two opposite parallel edges of the target certificate;

calculating a ratio between the two second length parameters, and judging whether the ratio is smaller than a second preset threshold value;

And if the ratio is smaller than a second preset threshold value, stopping shooting the target certificate to obtain the video stream.

The application also provides a certificate authenticity verification device, including:

the system comprises a characteristic detection module, a color changing module and a color changing module, wherein the characteristic detection module is used for acquiring image information of a target certificate at different continuous shooting angles, performing characteristic detection on the image information at each shooting angle and acquiring a plurality of continuous characteristic information, the characteristic detection comprises color changing ink value detection on an optical color changing ink area in the image information, and the characteristic information comprises a color changing ink value;

The calculation module is used for calculating the difference value between every two adjacent color-changing ink values and taking the absolute value to obtain a plurality of color-changing ink absolute values;

the first judgment module is used for judging whether a first color-changing ink absolute value larger than a preset color-changing ink absolute value exists in the color-changing ink absolute values or not;

And the output module is used for outputting the information that the target certificate is true when a first color-changing ink absolute value larger than a preset color-changing ink absolute value exists in the plurality of color-changing ink absolute values.

The application also provides a computer device, which comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the certificate authenticity verification method when executing the computer program.

The present application further provides a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the steps of the certificate authenticity verification method described above.

the beneficial effect of this application is: the certificate authenticity verification method provided by the application comprises the steps of obtaining image information of a target certificate at different continuous shooting angles, detecting color-changing ink values in optical color-changing ink areas in each image information, continuously obtaining a plurality of color-changing ink values, calculating the difference value between every two adjacent color-changing ink values, obtaining absolute values of the color-changing inks, and verifying authenticity of the target certificate by judging whether the absolute values of the color-changing inks are larger than the first absolute value of the preset color-changing inks or not.

Drawings

FIG. 1 is a schematic flow chart of a method for verifying authenticity of a document in one embodiment of the present application;

FIG. 2 is a schematic diagram of a document authentication device in one embodiment of the present application;

FIG. 3 is a schematic diagram of a computer device in an implementation of the present application.

The implementation, functional features and advantages of the objectives of the present application will be further explained with reference to the accompanying drawings.

Detailed Description

It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.

Referring to fig. 1, an embodiment of the present application provides a certificate authenticity verification method, which can be applied to a terminal device with a camera function (e.g., an intelligent robot with a camera function, etc.), and the certificate authenticity verification method includes:

s11, acquiring image information of the target certificate at different continuous shooting angles, and performing feature detection on the image information at each shooting angle to acquire a plurality of continuous feature information, wherein the feature detection comprises color-changing ink value detection, and the feature information comprises color-changing ink values;

s12, calculating the difference between every two adjacent color-changing ink values and taking the absolute value to obtain a plurality of absolute values of the color-changing ink;

S13, judging whether a first color-changing ink absolute value larger than a preset color-changing ink absolute value exists in the color-changing ink absolute values or not;

and if the first color-changing ink absolute value which is larger than the preset color-changing ink absolute value exists in the plurality of color-changing ink absolute values, executing S14, and outputting the information that the target certificate is true.

In the above S11, the target document is an identity document with an optical color-changing ink anti-counterfeit feature that needs to be authenticated currently, such as a second generation identity card in china continental, a permanent identity card in hong kong, and the like, wherein for the real second generation identity card in china continental, a great wall pattern with a size of about 22mm × 14mm printed by the optical color-changing ink is provided at a position with a distance of about 7.5mm from a left edge and about 9.5mm from an upper edge of the front face of the identity card, and the great wall pattern dynamically presents different colors along with a change of an observation angle, such as a color gradually changing from yellow to green or a color gradually changing from green to yellow; for the real hong Kong permanent resident ID card, a triangular pattern printed by optical color-changing ink is arranged at the left side of the wafer on the front surface of the ID card, and the triangular pattern can dynamically present different colors along with the change of the observation angle, such as the color is gradually changed from golden to green or from green to golden. In this step, specifically, the terminal device may continuously capture the front side of the target document from different capture angles to obtain a plurality of continuous document images corresponding to different angles, and then perform targeted clipping on the document images to clip out first area images containing optical color-changing ink anti-counterfeiting features from each document image, and further convert the clipped first area images from RGB color spaces to HSV color spaces through an opencv or other image processing library to obtain HSV values corresponding to each first area image, and then extract an H value (where the H value is an integer) of the HSV values to obtain the color-changing ink value, thereby implementing color-changing ink value detection on image information at each capture angle.

in the above S12, after obtaining a plurality of color-changing ink values, a plurality of color-changing ink absolute values may be obtained by calculating the difference between two adjacent color-changing ink values and taking the absolute value, for example, if the number of obtained color-changing ink values is 10, then 9 color-changing ink absolute values may be obtained by calculation.

in S13, specifically, since the ink area containing the security feature of the optically variable ink dynamically shows different colors with the change of the viewing angle for the real target document, for example, for the real second generation resident identification card of mainland china, the color gradually changes from yellow to green or from green to yellow, and for example, for the real permanent resident identification card of hong kong, the color gradually changes from gold to green or from green to gold, and the different colors correspond to different HSV values, for example, yellow corresponds to HSV value (60, 100, 100) (i.e. yellow corresponds to H value 60), gold corresponds to HSV value (51, 100, 100) (i.e. gold corresponds to H value 51), green corresponds to HSV value (120, 100, 50) (i.e. green corresponds to H value 120), so that there is a certain difference between two adjacent color-variable ink values, the difference can be represented by a numerical value interval, the numerical value interval can be obtained through a large number of experiments in advance, therefore, after a plurality of absolute values of the color-changing ink are obtained, whether the absolute value of the color-changing ink larger than the absolute value of the preset color-changing ink exists in the plurality of absolute values of the color-changing ink or not can be judged, and the authenticity of the target certificate can be verified.

in the above S14, if there is a first absolute value of the color-changing ink greater than the preset absolute value of the color-changing ink in the plurality of absolute values of the color-changing ink, it is indicated that the target document is a true identity card, and at this time, information that the target document is true may be output to prompt the user, otherwise, it is indicated that the target document is a false identity card, and at this time, information that the target document is false may be output to prompt the user.

in this embodiment, the certificate authenticity verification method obtains image information of a target certificate at different continuous shooting angles, detects color-changing ink values in optical color-changing ink areas in each image information, can continuously obtain a plurality of color-changing ink values, and further obtains absolute values by calculating a difference value between every two adjacent color-changing ink values and taking the absolute values, so as to obtain a plurality of color-changing ink absolute values capable of representing authenticity of the identity certificate, and further verifies authenticity of the target certificate by judging whether a first color-changing ink absolute value larger than a preset color-changing ink absolute value exists in the plurality of color-changing ink absolute values, so that authenticity verification of the identity certificate can be realized without depending on manual identification, and authenticity verification of the identity certificate is greatly improved in efficiency and accuracy of certificate authenticity verification.

In an optional embodiment, before the step of acquiring image information of the target document at different continuous shooting angles, and performing feature detection on the image information at each shooting angle to obtain a plurality of continuous feature information, the method further includes:

s101, receiving a selection instruction input by a user, and determining the certificate type of the target certificate according to the selection instruction;

S102, according to the certificate type, determining a preset color-changing ink absolute value corresponding to the certificate type from the preset certificate type-color-changing ink absolute value corresponding relation.

In this embodiment, in order to meet the diversity of business requirements, the authenticity verification of different types of target certificates can be implemented through the terminal device. Specifically, the user may determine the certificate type of the target certificate through a manual selection manner, for example, a certificate type option may be provided for the user on a specific application interface of the terminal device, for example, the certificate type option may be mainland china, hong kong china, and the like, and the user may select the certificate type of the target certificate through a click manner and the like, and when the user performs an operation such as a click on the option, the terminal device may be triggered to generate a corresponding selection instruction, and then the terminal device may determine the certificate type of the target certificate according to the selection instruction, for example, if the user clicks the certificate type option of "hong kong", it indicates that the current authentication required is the authenticity of the permanent resident identity card in hong kong; wherein, the certificate type-color-changing ink absolute value corresponding relation can be stored in the terminal equipment in advance in a form of a table, the certificate type-color-changing ink absolute value corresponding relation is pre-stored with color-changing ink absolute values corresponding to all target certificate types (generally, the color-changing ink absolute values of the identity card authenticity critical points are different for different certificate types), such as the absolute value of the color-changing ink corresponding to the second generation resident identification card of mainland china, the absolute value of the color-changing ink corresponding to the permanent resident identification card of hong kong, etc., and therefore, after the certificate type of the target certificate is determined, the terminal equipment searches the preset certificate type-color-changing ink absolute value corresponding relation, the preset color-changing ink absolute value corresponding to the certificate type of the target certificate can be determined, so that the authenticity of the identity card can be judged by utilizing the preset color-changing ink absolute value. In this embodiment, it should be noted that, in some specific usage scenarios, the user may also set the default certificate type by himself/herself, for example, for some usage scenarios mainly for verifying the authenticity of the second generation identity card in continental china, the user may set the default certificate type by himself/herself as continental china, so as to avoid frequent manual selection operations.

in another optional embodiment, the feature detection further includes performing character detection by using a preset character recognition neural network model, the feature information further includes a character detection result, the step of obtaining image information of the target document at different continuous shooting angles, performing feature detection on the image information at each shooting angle, and obtaining a plurality of continuous feature information further includes:

S11a, determining the certificate type of the target certificate according to the detection results of the characters;

s11b, according to the certificate type, determining the preset color-changing ink absolute value corresponding to the certificate type from the preset certificate type-color-changing ink absolute value corresponding relation.

In this embodiment, while detecting the color-changing ink value of the optical color-changing ink area in the image information at each shooting angle, the preset character recognition neural network model is used to perform character detection on the area (such as the optical character area) containing the specific character in the image information at each shooting angle, and then the certificate type of the target certificate can be automatically determined according to the result of the character detection, so that the trouble of manual operation can be avoided, and the efficiency of certificate authenticity verification can be further improved. For example, for a second generation resident identification card of the true continental CHINA, an anti-counterfeiting characteristic character of the 'CHINA' with the size of about 17.5mm × 4.0mm, which is produced by a micro-lens micro-pattern combined film technology, is arranged under a picture on the right side of the front surface of the identification card, and the anti-counterfeiting characteristic character can be dynamically presented along with the change of an observation angle; for the real hong Kong permanent resident identification card, an oblong anti-counterfeiting feature, namely a multiple laser image, is arranged at the lower side of the wafer on the front surface of the identification card, and the multiple laser image can dynamically present H, K characters along with the change of the observation angle. Specifically, in some embodiments, the terminal device may continuously capture the front side of the target document from different capturing angles to obtain a plurality of continuous document images corresponding to different angles, and then, by performing targeted cropping on the document images, a second area image (e.g., a second area image containing multiple laser images) containing the anti-counterfeiting character pattern may be cropped from each document image, and further, by inputting the cropped second area image into a preset character recognition neural network model for character detection, character detection results corresponding to each second area image may be obtained, so as to implement character detection on image information at each capturing angle, where the character detection results include a plurality of types, for convenience of description, the target document is a hong kong permanent identity card, and it is assumed that the target document is a hong kong permanent identity card (i.e., the document type is hong kong), then, by performing character detection on the optical character area (i.e. the area containing the multiple laser images) in the image information at each shooting angle, 4 character detection results can be obtained: contains only H characters, only K characters, neither H, K characters, and all H, K characters; further, in one mode, it is possible to determine whether the certificate type of the target certificate is hong Kong, China by judging whether there is a character detection result each containing H, K characters among the plurality of character detection results obtained, in another mode, whether the certificate type of the target certificate is hong Kong can be determined by judging whether at least one character detection result only containing H characters and at least one character detection result only containing K characters exist in the obtained multiple character detection results, and if the multiple character detection results exist character detection results all containing H, K characters or at least one character detection result only containing H characters and at least one character detection result only containing K characters exist in the multiple character detection results, the terminal equipment can determine that the certificate type of the target certificate is hong Kong; similarly, according to the detection results of multiple characters, it can also be determined whether the document type of the target document is mainland china, and therefore, after determining the document type of the target document, the terminal device can determine the preset absolute value of the color-changing ink corresponding to the document type of the target document by looking up the preset document type-absolute value of the color-changing ink corresponding relationship, so as to subsequently determine whether the identity card is true or false by using the preset absolute value of the color-changing ink, where it should be noted that, in this embodiment, in addition to performing character detection on the optical character region in the image information by using the preset character recognition neural network model, in other embodiments, the preset character recognition neural network model can also be used to perform character detection on other character regions in the image information, for example, for a permanent resident identity card in hong kong, the upper edge area of the front surface of the target certificate is printed with a 'hong Kong permanent resident identification card', so that the character detection can be carried out on the area, and when the character detection result contains a character detection result which can represent a 'hong Kong' two-character, the type of the target certificate can be determined to be hong Kong in China.

in addition, it should be noted that, in some other embodiments, the manual selection of the certificate type and the automatic determination of the certificate type according to the character detection result may be further used to specify the certificate type of the target certificate, so that it is avoided that the accuracy of certificate authenticity verification is affected by a manual operation mistake (for example, when an application scene is changed, for example, when the application scene is changed from continental china to hong kong district, the default set certificate type is not adaptively adjusted), and when the manually selected certificate type (or the default set certificate type) is detected to be inconsistent with the automatically detected certificate type, a prompt message indicating that the certificate type is inconsistent may be issued to remind the user of the relevant checking operation.

In an optional embodiment, the feature detection further includes performing character detection on an optical character region in the image information by using a preset character recognition neural network model, the feature information further includes a character detection result, and before the step of outputting the information that the target certificate is true, the method further includes:

s140a, judging whether a first character detection result meeting the preset character condition exists in the plurality of character detection results;

If the first character detection result meeting the preset character condition exists in the plurality of character detection results, the step S14 is executed to output the information that the target certificate is true.

In this embodiment, for convenience of description, the target document is taken as a hong kong permanent resident identification card as an example, and if the target document is the hong kong permanent resident identification card, 4 character detection results can be obtained by performing character detection on an optical character area (i.e. an area containing multiple laser images) in image information at each shooting angle: contains only H characters, only K characters, neither H, K characters, and all H, K characters; further, in one mode, whether a first character detection result meeting a preset character condition exists in the plurality of character detection results can be judged by judging whether character detection results all containing H, K characters exist in the plurality of obtained character detection results, and in another mode, whether a first character detection result meeting a preset character condition exists in the plurality of obtained character detection results can be judged by judging whether at least one character detection result only containing H characters exists in the plurality of obtained character detection results and at least one character detection result only containing K characters exists in the plurality of obtained character detection results; if the obtained character detection results only contain the character detection result of the H character or only contain the character detection result of the K character, the terminal equipment can output the information that the target certificate is false according to the character detection results; if the character detection results all contain H, K characters, or at least one character detection result only containing H characters and at least one character detection result only containing K characters exist in the character detection results, the terminal device can judge that the first character detection result meeting the preset character condition exists in the character detection results, and further can output the information that the target certificate is true, namely, the terminal device can output the information that the target certificate is true only when the character detection results all contain H, K characters exist, or at least one character detection result only containing H characters and at least one character detection result only containing K characters exist in the character detection results, and at least one first color-changing ink absolute value larger than the preset color-changing ink absolute value exists in the color-changing ink absolute values, therefore, the authenticity verification accuracy of the certificate can be further improved, and the authenticity verification of the identity certificate can be more reliable.

In an optional embodiment, before the step of obtaining image information of the target document at different continuous shooting angles, and performing feature detection on the image information at each shooting angle to obtain a plurality of continuous feature information, the method further includes:

S10a, acquiring a plurality of true certificate images with different inclination angles and carrying character marking information;

s10b, inputting a plurality of true certificate images into a character recognition neural network model for character detection, and outputting a plurality of character information;

s10c, comparing each character information with the corresponding character mark information one by one to verify whether the detection precision of the character recognition neural network model reaches the preset precision;

if the detection precision of the character recognition neural network model does not reach the preset precision, executing S10d, and repeatedly iterating the parameters in the character recognition neural network model until the detection precision of the character recognition neural network model reaches the preset precision.

in S10a, specifically, since the text is dynamically presented in the areas containing the anti-counterfeit character at different angles for the true identity document, a plurality of (e.g. 5, 10, 20, 50, etc.) same true identity document images at different angles of inclination can be obtained as training data, wherein the true identity document images can be from the user input or from the local, and are not limited to this; the character marking information is the character features of the standard typeface categories (such as the character features of the standard typeface categories of "H" and "K", the character features of the standard typeface categories of "middle", "country", "C", "H", "I", "N", "a", etc.), which can be manually marked in advance in the corresponding genuine certificate images for the subsequent use in calculating the accuracy of the training result, for example, if the genuine certificate images are from the image information collection of the genuine hong Kong permanent resident identification card, the character marking information of each genuine certificate image can be manually marked in advance, for example, the character marking information can be (1, 0) representing the character "H", (0, 1) representing the character "K", (0, 0) representing no character, (1, 1) representing the character "HK, it needs to be explained that, the above-mentioned true certificate image may be an area image containing an anti-counterfeit characteristic typeface in the identity certificate image, or an area image containing a specific character in the identity certificate image, which is not specifically limited, and for convenience of understanding, the embodiment takes the case where the true certificate image is an area image containing an anti-counterfeit characteristic typeface in the identity certificate image as an example to perform the relevant explanation.

in the above S10b, a classifier trained by a convolutional neural network may be used as the character recognition neural network model, and of course, other types of classifiers trained by a neural network may also be used as the character recognition neural network model as long as the usage requirement can be met, which is not particularly limited; specifically, when the character recognition neural network model receives an input genuine document image, the character recognition neural network model performs character detection on the inputted genuine document image, and can output corresponding character information, for example, when the output character information contains (1, 0), it represents that only an "H" character is detected in a certain genuine document image in the character detection process.

In the above S10c, the preset precision may be determined according to actual use requirements, and may be, for example, 80%, 85%, 90%, and the like, which is not particularly limited; for example, if there are 5 genuine document images, the preset precision is 85%, the character mark information corresponding to each genuine document image is (1, 0), (1, 1), (0, 1), (1, 0) and (0, 0), and the character information corresponding to each outputted genuine document image is (1, 0), (1, 1), (0, 0) and (0, 0), then comparing each character information with the corresponding character mark information one by one, so as to obtain a comparison result with four character information matches, and then calculating the detection precision of the character recognition neural network model to be 80% according to the comparison result, and then determining that the detection precision of the character recognition neural network model does not reach the preset precision.

in S10d, specifically, when the detection accuracy of the character recognition neural network model does not reach the preset accuracy, the prediction error between the output character information and the standard character marking information may be calculated in a loss function manner, and the prediction error is continuously reduced to the minimum by using a random gradient descent method, so that the character recognition neural network model with better network parameters may be trained.

In the embodiment, by acquiring a plurality of true certificate images with different inclination angles and carrying character marking information, inputting a plurality of true certificate images into a preset character recognition neural network model for character detection, outputting a plurality of character information, further comparing the output character information with the corresponding character mark information one by one, verifying whether the detection precision of the character recognition neural network model reaches the preset precision, when the detection precision of the character recognition neural network model does not reach the preset precision, the parameters (such as network parameters such as weight) in the character neural network model are iterated repeatedly until the detection precision of the character recognition neural network model reaches the preset precision, therefore, the character recognition neural network model with better network parameters can be trained, and the reliability of certificate authenticity verification can be further improved.

In an optional embodiment, the characteristic detection further includes performing color detection on an optically variable ink region in the image information by using a preset color recognition neural network model, the characteristic information further includes a color detection result, and before the step of outputting the information that the target certificate is true, the method further includes:

S140A, judging whether a first color detection result meeting a preset color condition exists in the multiple color detection results;

if there is a first color detection result meeting the preset color condition in the plurality of character detection results, the above S14 is executed to output information that the target certificate is true.

In this embodiment, for convenience of description, the target document is taken as a permanent resident identification card in hong Kong as an example, because for the real permanent resident identification card in hong Kong, a triangular pattern printed by the photochromic ink is disposed at the left side of the wafer on the front side of the identification card, and the triangular pattern dynamically presents different colors with the change of the observation angle, such as the color gradually changes from gold to green or from green to gold, in the specific implementation, the front side of the target document can be continuously photographed from different photographing angles to obtain a plurality of continuous document images corresponding to different angles, then by performing targeted cutting on the document images, the first region image containing the anti-counterfeiting feature of the photochromic ink can be cut out from each document image, and further, by inputting the cut first region image into a preset color recognition neural network model for color detection, obtaining color detection results corresponding to the first area images, so that color detection of image information at each shooting angle is realized, and further whether a detection result with a yellow color and a detection result with a green color exist in the multiple color detection results is judged; if there is no yellow or green detection result in the multiple color detection results, the terminal device can output false information of the target certificate, and if there is at least one yellow detection result and at least one green detection result in the multiple color detection results, the terminal device can determine that there is a first color detection result meeting the preset color condition in the multiple color detection results, and further can output true information of the target certificate, that is, only when there is at least one yellow detection result and at least one green detection result in the multiple color detection results and there is at least one first absolute value of color ink greater than the preset absolute value of color ink in the multiple absolute values of color ink, the terminal device can output true information of the target certificate, therefore, the authenticity verification accuracy of the certificate can be further improved, and the authenticity verification of the identity certificate can be more reliable. In this embodiment, it should be noted that the preset color recognition neural network model may be a pre-trained convolutional neural network model with a color recognition function, may also be a pre-trained BP neural network model with a color recognition function, and may also be other types of color classifiers as long as the usage requirements are met, and no specific limitation is imposed on the color recognition neural network model, taking the target certificate as a hong kong permanent resident identification card as an example, a color recognition neural network model (equivalent to a color two classifier capable of classifying gold and green) capable of recognizing gold and green may be trained in advance based on a convolutional neural network, and the training process is similar to the training process of the character recognition neural network model, and those skilled in the art can understand that details are not repeated here.

in addition, in some other embodiments, the color-changing ink value detection, the character detection and the color detection can be simultaneously performed on the image information at each shooting angle, and only when the color-changing ink value detection, the character detection and the color detection all pass (that is, at least one first color-changing ink absolute value greater than the preset color-changing ink absolute value exists in the plurality of color-changing ink absolute values, and a first character detection result meeting the preset character condition exists in the plurality of character detection results, and a first color detection result meeting the preset color condition exists in the plurality of color detection results), the terminal device can output the information that the target certificate is true, so that the accuracy of the authenticity verification of the identity certificate can be further improved, and the authenticity verification of the identity certificate can be more reliable.

In an optional embodiment, when the image information is a video stream, acquiring image information of the target certificate at different continuous shooting angles, and performing feature detection on the image information at each shooting angle to obtain a plurality of continuous feature information, including:

s111, continuously shooting the front side of the target certificate from different shooting angles to obtain a video stream;

And S112, carrying out feature detection on the video stream frame by frame or interval preset frame number to obtain a plurality of continuous feature information.

in this embodiment, the terminal device may continuously shoot the front side of the target document for a predetermined time (e.g., 2 seconds, 3 seconds, 4 seconds, etc.) from different shooting angles, so as to obtain a video stream with a certain time length, and after obtaining the video stream, may perform feature detection (such as color change ink value detection and character detection) on the video stream in a frame-by-frame (i.e., frame-by-frame) manner, or may perform feature detection on the video stream in a frame-by-frame manner, where the preset frame number may be determined according to actual usage, such as 1 frame, 2 frames, 3 frames, etc., and is not specifically limited to this, and may meet usage requirements, and preferably, may perform feature detection on the video stream in a frame-by-frame manner, so that detection accuracy is not affected, and detection efficiency may be improved; specifically, after a video stream is obtained, by performing targeted cropping on a video frame (i.e., a frame image in the video stream), the first area image containing the anti-counterfeiting feature of the optical color-changing ink and the second area image containing the anti-counterfeiting feature character can be respectively cropped from each video frame (e.g., the 1 st frame, the 3 rd frame, the 5 th frame, the 7 th frame, etc.), wherein for the first area image, the cropped first area image can be converted from an RGB color space to an HSV color space through an opencv or other picture processing library, HSV values corresponding to each first area image can be obtained, and then an H value in the HSV values is extracted as the color-changing ink value, so that detection of the color-changing ink value of the certificate target is realized; and for the second area images, character detection can be performed by inputting the second area images into the character recognition neural network model, so that character detection results corresponding to each second area image can be obtained, and thus character detection of the target document can be achieved, therefore, in some embodiments, by performing color changing ink value detection and character detection on the video stream every other preset frame number, a corresponding color changing ink value and character detection result can be obtained, for example, assuming that the preset frame number is 3 frames, and the time length and the frame number of the video stream are 5 seconds and 30 frames/second respectively, by performing color changing ink value detection and character detection on the video stream in a manner of separating the preset frame numbers, 50 continuous color changing ink values and 50 continuous character detection results can be theoretically obtained. In this embodiment, the authenticity verification of the identification card is performed in a dynamic manner such as video streaming (i.e., video streaming), and compared with a technical means for performing the authenticity verification of the identification card in a static manner such as image information (document image), the reliability of the authenticity verification of the document is improved advantageously because the continuity and stability of information acquisition are higher.

In an alternative embodiment, the step of capturing the front side of the target document continuously from different capture angles to obtain the video stream comprises:

s1111, focusing the target certificate, and acquiring a first shape parameter of the focused target certificate, wherein the first shape parameter comprises first length parameters of two opposite edges of the target certificate;

s1112, calculating a deviation value between the two first length parameters, and determining whether the deviation value is smaller than a first preset threshold;

If the deviation value is smaller than the first preset threshold value, executing S1113, starting to continuously shoot the front side of the target certificate from different shooting angles, and acquiring a second shape parameter of the target certificate in real time in the shooting process, wherein the second shape parameter comprises second length parameters of two opposite parallel edges of the target certificate;

s1114, calculating a ratio between the two second length parameters, and judging whether the ratio is smaller than a second preset threshold value;

And if the ratio is smaller than a second preset threshold value, executing S1115, and stopping shooting the target certificate to obtain the video stream.

in S1111, specifically, at the beginning, focusing is performed on the front side of the target document by using the camera, and then the edge of the target document is detected by the edge detection technology, so as to obtain 4 first length parameters (i.e. lengths of 4 edges) corresponding to 4 edges of the target document respectively.

in the above step S1112, after obtaining 4 first length parameters corresponding to 4 edges of the target document, further calculating a length difference (i.e. a deviation value) between two opposite edges of the target document, and determining whether the length difference is smaller than a first preset threshold, if the length difference is smaller than the first preset threshold, it indicates that the lengths of the two opposite edges are approximately equal, the focused target document appears as a rectangle, and the camera is aligned with the front side of the target document (which is equivalent to that the camera is aligned with the front side of the target document).

in S1113, in the shooting process, the shape of the detected target certificate changes with the change of the shooting angle, and is specifically changed from a rectangle to a trapezoid with a short upper side and a long lower side, and 4 second length parameters corresponding to 4 edges when the target certificate is in the trapezoid can be detected in real time by using an edge detection technology in the shooting process.

in the above S1114, after obtaining 4 second length parameters of 4 sides when the target document is in the shape of a trapezoid, further calculating a length ratio between two opposite parallel edges (i.e. a ratio between an upper side length and a lower side length of the trapezoid), and determining whether the length ratio is smaller than a second preset threshold, where the second preset threshold may be obtained in advance through experiments, and for example, may be 0.73, 0.72, 0.71, etc., as long as the usage requirement can be met, which is not limited specifically; if the length ratio is smaller than a second preset threshold value, the target certificate is shown to be in a required trapezoidal shape, and enough 'materials' for certificate authenticity verification are collected by the camera.

In the above S1115, when it is detected that the length ratio between the two opposite parallel edges is smaller than the second preset threshold, it indicates that the camera has collected enough "material" for certificate authenticity verification, and it is not necessary to capture the target certificate, so that the camera can be controlled to stop capturing the target certificate at this time, and the video stream is obtained.

In this embodiment, as for a real target document, as the angle changes, the ink area containing the anti-counterfeiting feature of the optically variable ink dynamically presents different colors, and the area containing the word of the anti-counterfeiting feature dynamically presents the word, in order to verify the authenticity of the target document, it is necessary to "observe" the target document from different angles by simulating the actions of a human, wherein the different shooting angles not only affect the verification accuracy, but also the initial shooting angle and the final shooting angle finally determine the length of the video stream, thereby finally affecting the verification efficiency, so it is necessary to determine the initial angle and the final shooting angle for ensuring the verification accuracy and efficiency at the same time, and in this example, the angle facing the target document is selected as the initial angle for shooting, thereby simplifying the operation, and, the edge detection technology is used for carrying out edge detection on the target certificate, so that the automatic control of the shooting angle of the camera is realized, the shooting angle of the camera is not required to be changed by manual operation, and a proper termination angle can be finally obtained, so that the accuracy and the efficiency of verification can be ensured.

In an optional embodiment, the step of determining whether there is a first color-changing ink absolute value greater than a preset color-changing ink absolute value in the plurality of color-changing ink absolute values includes:

s131, selecting a color-changing ink absolute value with the largest numerical value from the plurality of color-changing ink absolute values;

s132, judging whether the absolute value of the color-changing ink with the largest numerical value is larger than the preset absolute value of the color-changing ink or not;

If the absolute value of the color-changing ink with the largest numerical value is larger than the preset absolute value of the color-changing ink, S133 is executed to determine that a first absolute value of the color-changing ink larger than the preset absolute value of the color-changing ink exists in the plurality of absolute values of the color-changing ink.

In this embodiment, as long as there is a color-changing ink absolute value greater than the preset color-changing ink absolute value in the plurality of color-changing ink absolute values, the target certificate can be determined to be true accordingly, and therefore, only the color-changing ink absolute value with the largest numerical value needs to be selected from the plurality of color-changing ink absolute values, and then, whether the target certificate is true can be determined by judging whether the color-changing ink absolute value with the largest numerical value is greater than the preset color-changing ink absolute value, without comparing the sizes of the plurality of color-changing ink values with the preset color-changing ink absolute value one by one, thereby being beneficial to shortening the operation process and improving the verification efficiency.

referring to fig. 2, the embodiment of the present application further provides a certificate authenticity verification apparatus, which can be applied to a terminal device with a camera shooting function, and the certificate authenticity verification apparatus includes:

The characteristic detection module 11 is configured to acquire image information of the target certificate at different continuous shooting angles, perform characteristic detection on the image information at each shooting angle, and acquire a plurality of continuous characteristic information, where the characteristic detection includes performing color-changing ink value detection on an optical color-changing ink area in the image information, and the characteristic information includes a color-changing ink value;

the calculation module 12 is used for calculating the difference value between every two adjacent color-changing ink values and taking the absolute value to obtain a plurality of color-changing ink absolute values;

The first judgment module 13 is configured to judge whether a first color-changing ink absolute value larger than a preset color-changing ink absolute value exists in the plurality of color-changing ink absolute values;

and the output module 14 is configured to output information that the target certificate is true when a first color-changing ink absolute value greater than a preset color-changing ink absolute value exists in the plurality of color-changing ink absolute values.

in an optional embodiment, the apparatus for verifying authenticity of a document further includes:

the receiving module is used for receiving a selection instruction input by a user and determining the certificate type of the target certificate according to the selection instruction;

The first determining module is used for determining a preset color-changing ink absolute value corresponding to the certificate type from a preset certificate type-color-changing ink absolute value corresponding relation according to the certificate type.

In an optional embodiment, the feature detection further includes performing character detection using a preset character recognition neural network model, the feature information further includes a character detection result, and the certificate authenticity verification apparatus further includes:

The second determining module is used for determining the certificate type of the target certificate according to the detection results of the plurality of characters;

and the third determining module is used for determining a preset color-changing ink absolute value corresponding to the certificate type from the preset certificate type-color-changing ink absolute value corresponding relation according to the certificate type.

In an optional embodiment, the feature detection further includes performing character detection on an optical character region in the image information by using a preset character recognition neural network model, and the certificate authenticity verification apparatus further includes:

and the second judgment module is used for judging whether a first character detection result meeting the preset character condition exists in the plurality of character detection results.

The output module 14 is further configured to output information that the target certificate is true when a first character detection result meeting a preset character condition exists in the plurality of character detection results.

In an optional embodiment, the characteristic detection further includes performing color detection on an optically variable ink region in the image information by using a preset color recognition neural network model, the characteristic information further includes a color detection result, and the certificate authenticity verification apparatus further includes:

And the third judging module is used for judging whether a first color detection result meeting the preset color condition exists in the plurality of color detection results.

The output module 14 is further configured to output information that the target certificate is true when a first color detection result meeting a preset color condition exists in the plurality of character detection results.

The acquisition module acquires a plurality of true certificate images with different inclination angles and carrying character marking information;

the character detection module is used for inputting a plurality of true certificate images into the character recognition neural network model for character detection and outputting a plurality of character information;

The comparison module is used for comparing each piece of character information with the corresponding character mark information one by one so as to verify whether the detection precision of the character recognition neural network model reaches the preset precision;

And the iteration module is used for repeatedly iterating parameters in the character recognition neural network model until the detection precision of the character recognition neural network model reaches the preset precision when the detection precision of the character recognition neural network model does not reach the preset precision.

in an optional embodiment, when the image information is a video stream, the feature detection module 11 includes:

the shooting unit is used for continuously shooting the front side of the target certificate from different shooting angles to obtain a video stream;

the characteristic detection unit is used for carrying out characteristic detection on the video stream frame by frame or interval preset frame number to obtain a plurality of continuous characteristic information;

in an optional embodiment, the shooting unit includes:

the focusing subunit is used for focusing the target certificate and acquiring a first shape parameter of the focused target certificate, wherein the first shape parameter comprises first length parameters of two opposite edges of the target certificate;

the first calculating subunit is used for calculating a deviation value between the two first length parameters and judging whether the deviation value is smaller than a first preset threshold value or not;

the shooting subunit is used for starting to continuously shoot the front side of the target certificate from different shooting angles when the deviation value is smaller than a first preset threshold value, and acquiring a second shape parameter of the target certificate in real time in the shooting process, wherein the second shape parameter comprises second length parameters of two opposite parallel edges of the target certificate;

the second calculating subunit is used for calculating a ratio between the two second length parameters and judging whether the ratio is smaller than a second preset threshold value;

and the stopping subunit is used for stopping shooting the target certificate when the ratio is smaller than a second preset threshold value to obtain the video stream.

in an optional embodiment, the first determining module 13 includes:

the selecting unit is used for selecting the color-changing ink absolute value with the largest value from the plurality of color-changing ink absolute values;

the judgment unit is used for judging whether the absolute value of the color-changing ink with the largest numerical value is larger than the preset absolute value of the color-changing ink or not;

And the judging unit is used for judging that a first color-changing ink absolute value larger than the preset color-changing ink absolute value exists in the plurality of color-changing ink absolute values when the color-changing ink absolute value with the largest value is larger than the preset color-changing ink absolute value.

With regard to the certificate authenticity verifying apparatus in the above embodiments, the specific manner in which each module performs operations has been described in detail in the embodiments related to the method, and is not described herein again.

referring to fig. 3, a computer device, which may be a server and whose internal structure may be as shown in fig. 3, is also provided in the embodiment of the present application. The computer device includes a processor, a memory, a network interface, and a database connected by a system bus. Wherein the computer designed processor is used to provide computational and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used for storing certificate authenticity verification method programs and the like. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is used for realizing the certificate authenticity verification method in any embodiment when being executed by a processor.

The embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the method for verifying authenticity of a certificate in any of the above embodiments is implemented.

it will be understood by those skilled in the art that all or part of the processes of the methods of the above embodiments may be implemented by hardware associated with instructions of a computer program, which may be stored on a non-volatile computer-readable storage medium, and when executed, may include processes of the above embodiments of the methods. Any reference to memory, storage, database, or other medium provided herein and used in the examples may include non-volatile and/or volatile memory. Non-volatile memory can include read-only memory (ROM), Programmable ROM (PROM), Electrically Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), double data rate SDRAM (SSRDRAM), Enhanced SDRAM (ESDRAM), synchronous Link (Synchlink) DRAM (SLDRAM), Rambus (Rambus) direct RAM (RDRAM), direct bused dynamic RAM (DRDRAM), and bused dynamic RAM (RDRAM).

It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, apparatus, article, or method. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, apparatus, article, or method that includes the element.

the above description is only a preferred embodiment of the present application, and not intended to limit the scope of the present application, and all modifications of equivalent structures and equivalent processes, which are made by the contents of the specification and the drawings of the present application, or which are directly or indirectly applied to other related technical fields, are also included in the scope of the present application.

Claims

1. A method for verifying authenticity of a certificate, comprising:

acquiring image information of a target certificate at different continuous shooting angles, and performing feature detection on the image information at each shooting angle to acquire a plurality of continuous feature information, wherein the feature detection comprises color-changing ink value detection on an optical color-changing ink area in the image information, and the feature information comprises color-changing ink values;

Calculating the difference value between every two adjacent color-changing ink values and taking an absolute value to obtain a plurality of color-changing ink absolute values;

and if a first color-changing ink absolute value which is larger than a preset color-changing ink absolute value exists in the plurality of color-changing ink absolute values, outputting the information that the target certificate is true.

2. the method for verifying authenticity of a document according to claim 1, wherein the feature detection further comprises character detection using a preset character recognition neural network model, the feature information further comprises a character detection result, the steps of obtaining image information of the target document at different continuous shooting angles, performing the feature detection on the image information at each shooting angle, and obtaining a plurality of continuous feature information further comprise:

Determining the certificate type of the target certificate according to a plurality of character detection results;

And determining the preset color-changing ink absolute value corresponding to the certificate type from the preset certificate type-color-changing ink absolute value corresponding relation according to the certificate type.

3. the method for verifying authenticity of a document according to claim 1, wherein the feature detection further includes character detection of an optical character area in the image information using a preset character recognition neural network model, the feature information further includes a character detection result, and before the step of outputting the information that the target document is genuine, the method further includes:

and if a first character detection result meeting a preset character condition exists in the plurality of character detection results, executing the step of outputting the information that the target certificate is true.

4. the method for verifying authenticity of a document according to claim 1, wherein the detecting of the characteristics further comprises detecting colors of the optically variable ink areas in the image information by using a preset color recognition neural network model, the characteristic information further comprises a color detection result, and before the outputting of the information that the target document is genuine, the method further comprises:

judging whether a first color detection result meeting a preset color condition exists in the plurality of color detection results;

And if a first color detection result meeting a preset color condition exists in the plurality of character detection results, executing the step of outputting the information that the target certificate is true.

5. The method for verifying authenticity of a document according to claim 3, wherein before the step of obtaining image information of the target document at different continuous shooting angles and performing feature detection on the image information at each shooting angle to obtain a plurality of continuous feature information, the method further comprises:

inputting a plurality of true certificate images into the character recognition neural network model for character detection, and outputting a plurality of character information;

comparing each piece of character information with the corresponding character mark information one by one to verify whether the detection precision of the character recognition neural network model reaches the preset precision or not;

if the detection precision of the character recognition neural network model does not reach the preset precision, repeatedly iterating parameters in the character recognition neural network model until the detection precision of the character recognition neural network model reaches the preset precision.

6. the method for verifying authenticity of a document according to claim 1, wherein when the image information is a video stream, the step of obtaining image information of the target document at different continuous shooting angles and performing feature detection on the image information at each shooting angle to obtain a plurality of continuous feature information comprises:

continuously shooting the front side of the target certificate from different shooting angles to obtain the video stream;

and carrying out the characteristic detection on the video stream frame by frame or interval preset frame number to obtain a plurality of continuous characteristic information.

7. the method of claim 6, wherein the step of continuously capturing the front side of the target document from different capturing angles to obtain the video stream comprises:

8. a device for verifying authenticity of a document, comprising:

The system comprises a characteristic detection module, a color changing module and a color changing module, wherein the characteristic detection module is used for acquiring image information of a target certificate at different continuous shooting angles, performing characteristic detection on the image information at each shooting angle to acquire a plurality of continuous characteristic information, the characteristic detection comprises color changing ink value detection on an optical color changing ink area in the image information, and the characteristic information comprises a color changing ink value;

9. a computer device comprising a memory and a processor, the memory storing a computer program, wherein the processor when executing the computer program implements the steps of the document authenticity verification method according to any one of claims 1 to 7.

10. a computer-readable storage medium, having a computer program stored thereon, the computer program, when being executed by a processor, performing the steps of the method of verifying the authenticity of a document as claimed in any one of claims 1 to 7.