Iris registration method, iris recognition method, computer equipment and storage medium

Abstract

The present application provides an iris registration method, an iris recognition method, a computer device and a storage medium, the iris registration method comprising: acquiring an iris image combination corresponding to a target user, the iris image combination comprising a plurality of iris images of the target user, the plurality of iris images being iris images corresponding to different angles; calculating three-dimensional iris point cloud data based on the iris image combination, the three-dimensional iris point cloud data being used for reflecting iris features of the target user and spatial positions of the iris features; and according to the three-dimensional iris point cloud data, generating a target three-dimensional iris model of the target user, the target three-dimensional iris model being used for reflecting iris curved surface features of the target user. According to the technical scheme, the security of the iris recognition technology can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of iris recognition, and in particular to an iris registration method, an iris recognition method, a computer device and a storage medium. BACKGROUND

[0002] The iris is a colored ring of tissue around the pupil of the human eye, and its texture structure is randomly formed during embryonic development, with high stability and uniqueness. Even for identical twins or the eyes of the same person, the iris features are completely different. Therefore, iris recognition technology is used for identity verification in the field of biometric features.

[0003] Generally, iris recognition technology collects iris images and extracts the two-dimensional texture structure (such as wrinkles, spots or stripes, etc.) of the iris in the iris images as iris features, and compares them with the iris feature templates previously recorded and stored in the iris database to verify the identity. Two-dimensional iris features are easy to obtain by fraudulent means, such as stealing or forging iris images to pass the verification, which poses a risk of being attacked by counterfeiting. SUMMARY

[0004] The present application provides an iris registration method, an iris recognition method, a computer device and a storage medium, aiming to solve the technical problem of the risk of counterfeiting attack in iris recognition technology.

[0005] In a first aspect, an iris registration method is provided, comprising: obtaining an iris image combination corresponding to a target user, the iris image combination comprising a plurality of iris images of the target user, the plurality of iris images being iris images corresponding to different angles; based on the iris image combination, calculating three-dimensional iris point cloud data, the three-dimensional iris point cloud data being used to reflect the iris features of the target user and the spatial positions of the iris features; generating a target three-dimensional iris model of the target user according to the three-dimensional iris point cloud data, the target three-dimensional iris model being used to reflect the iris surface features of the target user.

[0006] In the technical solution, a target user's iris image combination is acquired, the iris image combination includes iris images corresponding to different angles of the target user, then three-dimensional iris point cloud data is calculated based on the iris image combination, the three-dimensional iris point cloud data is used to reflect iris features and spatial positions of the iris features of the target user, and finally a target three-dimensional iris model of the target user is generated according to the three-dimensional iris point cloud data, the target three-dimensional iris model is used to reflect iris curved surface features of the target user; since the target three-dimensional iris model of the user is generated in advance, the target three-dimensional iris model reflects the iris curved surface features of the user, compared with two-dimensional iris features, the iris curved surface features also cover depth information of the iris features, and thus the iris features obtained by means of a fake sample (such as a photo or contact lens projection) cannot pass the verification when compared with the target three-dimensional iris model due to the lack of the depth information of the iris features, so that the security of the iris recognition technology is improved.

[0007] With reference to the first aspect, in a possible implementation manner, the acquiring the iris image combination corresponding to the target user includes: acquiring a plurality of iris image sets corresponding to the target user, the plurality of iris image sets are iris image sets corresponding to a plurality of angles, and each iris image set includes a plurality of iris images of the target user at a same angle; performing image fusion on the plurality of iris images in a same iris image set to obtain the iris image combination, and the plurality of iris images in the iris image combination include images obtained by performing image fusion on the plurality of iris image sets respectively.

[0008] For each angle, a plurality of iris images are acquired to perform image fusion to obtain an iris image corresponding to each angle to form the iris image combination, so that the quality of the iris image is improved, and a higher-resolution iris image is obtained.

[0009] With reference to the first aspect, in a possible implementation manner, the iris image combination includes an iris image corresponding to a target angle, and the target angle is one of the plurality of angles; the performing image fusion on the plurality of iris images in a same iris image set to obtain the iris image combination includes: calculating an optical flow vector between two adjacent iris images in a target iris image set, the target iris image set is an iris image set corresponding to the target angle; performing registration on each iris image in the target iris image set based on the optical flow vector; and performing superimposed fusion on the each iris image after registration to obtain the iris image corresponding to the target angle.

[0010] The same angle corresponding to each frame of iris image is registered based on the optical flow vector between two adjacent frames of iris images, and the registered iris images are superimposed and fused, which can ensure that homologous pixels (such as the same object point) are accurately superimposed during fusion, reduce artifacts, and improve the image quality of the fused iris image.

[0011] With reference to the first aspect, in a possible implementation manner, before the image fusion is performed on the plurality of frames of iris images in the target iris image set, the method further includes: in a case where there is an iris image in the target iris image set that does not satisfy the preset quality condition, reacquiring an iris image of the target user at the target angle until all the iris images in the target iris image set satisfy the preset quality condition.

[0012] In a case where the iris image does not satisfy the preset quality condition, the iris image is reacquired, which can ensure the quality of the iris image.

[0013] With reference to the first aspect, in a possible implementation manner, the calculating three-dimensional iris point cloud data based on the iris image combination includes: performing iris feature extraction on each iris image in the iris image combination to obtain iris feature points corresponding to the plurality of iris images; performing feature matching on the iris feature points corresponding to the plurality of iris images to obtain a plurality of groups of iris feature matching pairs, each group of iris feature matching pairs including two iris feature points that match each other, the two iris feature points that match each other being iris feature points corresponding to each other in two iris images in the iris image combination; and calculating three-dimensional iris point cloud data based on the plurality of groups of iris feature point pairs.

[0014] The three-dimensional iris point cloud data is calculated based on iris feature matching, which is less affected by light and is helpful to accurately determine the spatial position of the iris feature.

[0015] With reference to the first aspect, in a possible implementation manner, the generating a target three-dimensional iris model of the target user according to the three-dimensional iris point cloud data includes: performing surface reconstruction on the three-dimensional iris point cloud data to obtain the target three-dimensional iris model of the target user.

[0016] The target three-dimensional iris model is obtained through surface reconstruction, so that the target three-dimensional iris model can reflect the iris surface characteristics of the user, thereby covering the depth information of the iris feature.

[0017] In a possible implementation manner of the first aspect, after the target three-dimensional iris model of the target user is generated according to the three-dimensional iris point cloud data, the method further includes: encrypting the target three-dimensional iris model to obtain an iris encryption template corresponding to the target user; segmenting the iris encryption template to obtain a plurality of iris encryption template segments corresponding to the target user; and storing different iris encryption template segments in the plurality of iris encryption template segments in different preset storage devices.

[0018] By encrypting the target three-dimensional iris model and segmenting the target three-dimensional iris model into a plurality of iris encryption template segments for distributed storage, leakage of the target three-dimensional iris model is prevented, and security of the iris recognition technology is improved.

[0019] In a possible implementation manner of the first aspect, after the target three-dimensional iris model of the target user is generated according to the three-dimensional iris point cloud data, the method further includes: encrypting the target three-dimensional iris model to obtain an iris encryption template corresponding to the target user; segmenting the iris encryption template to obtain a plurality of iris encryption template segments corresponding to the target user; and storing different iris encryption template segments in the plurality of iris encryption template segments in different preset storage devices.

[0020] By segmenting the key corresponding to the iris encryption template into a plurality of key segments for distributed storage, decryption difficulty of the iris encryption template is increased, and leakage of the target three-dimensional iris model is further prevented.

[0021] The second aspect provides an iris recognition method, including: obtaining a current three-dimensional iris model of a target user, the current three-dimensional iris model being used to reflect iris surface characteristics of the target user; obtaining a target three-dimensional iris model of the target user, the target three-dimensional iris model of the target user being generated in advance by the method of the first aspect; comparing the current three-dimensional iris model with the target three-dimensional iris model to obtain an iris recognition result of the target user.

[0022] In the technical solution, after obtaining the current three-dimensional iris model of the target user, the target three-dimensional iris model of the target user is obtained, and the current three-dimensional iris model and the target three-dimensional iris model are compared to obtain the iris recognition result of the target user; since the target three-dimensional iris model of the user is generated in advance, the target three-dimensional iris model reflects the iris surface feature of the user, and compared with the two-dimensional iris feature, the depth information of the iris feature is also covered in the iris surface feature; the iris feature obtained by means of a fake sample (such as a photo, a contact lens projection, etc.) cannot pass the verification when compared with the target three-dimensional iris model due to the lack of depth information of the iris feature, so that the security of the iris recognition technology is improved.

[0023] With reference to the second aspect, in a possible implementation manner, the obtaining of the target three-dimensional iris model of the target user comprises: obtaining, from different preset storage devices, iris encryption template fragments corresponding to the target user to obtain a plurality of iris encryption template fragments corresponding to the target user; combining the plurality of iris encryption template fragments to obtain an iris encryption template corresponding to the target user; and decrypting the iris encryption template to obtain the target three-dimensional iris model of the target user.

[0024] The distributed storage of the iris encryption template fragments helps to prevent leakage of the target three-dimensional iris model and improve security.

[0025] With reference to the second aspect, in a possible implementation manner, the decrypting of the iris encryption template to obtain the target three-dimensional iris model of the target user comprises: obtaining, from the different preset storage devices, key fragments corresponding to the target user to obtain a plurality of key fragments corresponding to the target user; combining the plurality of key fragments to obtain a key corresponding to the iris encryption template; and decrypting the iris encryption template by using the key to obtain the target three-dimensional iris model of the target user.

[0026] The distributed storage of the key fragments corresponding to the iris encryption template increases the decryption difficulty of the iris encryption template and further prevents leakage of the target three-dimensional iris model.

[0027] With reference to the second aspect, in a possible implementation manner, the obtaining of the current three-dimensional iris model of the target user comprises: obtaining a plurality of iris images of the target user; calculating current three-dimensional iris point cloud data based on the plurality of iris images, the current three-dimensional iris point cloud data being used to reflect iris features of the target user and spatial positions of the iris features; and generating the current three-dimensional iris model of the target user according to the current three-dimensional iris point cloud data.

[0028] With reference to the second aspect, in a possible implementation manner, before the plurality of iris images of the target user are acquired, the method further includes: adjusting a light parameter of an iris image acquisition device according to an ambient light intensity.

[0029] Adjusting the light parameter of the iris image acquisition device according to the ambient light intensity helps to improve the quality of the iris images.

[0030] In a third aspect, an iris registration device is provided, including: an iris image acquisition module, configured to acquire an iris image combination corresponding to a target user, the iris image combination including a plurality of iris images of the target user, the plurality of iris images being iris images corresponding to different angles; an iris point cloud calculation module, configured to calculate three-dimensional iris point cloud data based on the iris image combination, the three-dimensional iris point cloud data being used to reflect iris features of the target user and spatial positions of the iris features; an iris template generation module, configured to generate a target three-dimensional iris model of the target user according to the three-dimensional iris point cloud data, the target three-dimensional iris model being used to reflect iris surface features of the target user.

[0031] In a fourth aspect, an iris recognition device is provided, including: an iris model acquisition module, configured to acquire a current three-dimensional iris model of a target user, the current three-dimensional iris model being used to reflect iris surface features of the target user; an iris template acquisition module, configured to acquire a target three-dimensional iris model of the target user, the target three-dimensional iris model of the target user being generated in advance by the method of the first aspect; an iris recognition module, configured to compare the current three-dimensional iris model with the target three-dimensional iris model to obtain an iris recognition result of the target user.

[0032] In a fifth aspect, a computer device is provided, including a memory and a processor, the memory being connected to the processor, the processor being used to execute one or more computer programs stored in the memory, and the processor, when executing the one or more computer programs, causes the computer device to perform the iris registration method of the first aspect or the iris recognition method of the second aspect.

[0033] In a sixth aspect, a computer readable storage medium is provided, the computer readable storage medium storing a computer program, the computer program including program instructions, and the program instructions, when executed by a processor, causing the processor to perform the iris registration method of the first aspect or the iris recognition method of the second aspect.

[0034] The application can achieve the following technical effects: since the target three-dimensional iris model of the user is generated in advance, the target three-dimensional iris model reflects the iris surface characteristics of the user, and compared with two-dimensional iris characteristics, the iris surface characteristics also cover the depth information of the iris characteristics, and the iris characteristics obtained through the way of a fake sample (such as a photo, contact lens projection, etc.) cannot pass the verification when compared with the target three-dimensional iris model due to the lack of depth information of the iris characteristics, so that the security of the iris recognition technology can be improved. BRIEF DESCRIPTION OF DRAWINGS

[0035] In order to more clearly illustrate the technical solutions of the embodiments of the application, the following will briefly introduce the drawings needed to be used in the description of the embodiments of the application. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0036] Figure 1 A flowchart of an iris registration method provided by an embodiment of the application; Figure 2 A flowchart of another iris registration method provided by an embodiment of the application; Figure 3 A flowchart of an iris recognition method provided by an embodiment of the application; Figure 4 A structural diagram of an iris registration device provided by an embodiment of the application; Figure 5 A structural diagram of an iris recognition device provided by an embodiment of the application; Figure 6 A structural diagram of a computer device provided by an embodiment of the application. DETAILED DESCRIPTION

[0037] In order to make the purpose, technical solutions and advantages of the application more clear, the following will further describe the application in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the application, and are not used to limit the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application.

[0038] It should be noted that the various features of the embodiments of the present application can be combined with each other, and are within the protection scope of the present application, if there is no conflict. In addition, although the functional modules are divided in the device schematic diagram, and the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order from the module division in the device or the order in the flowchart. Furthermore, the "first", "second", "third" and the like used in the present application do not limit the data and execution order, but only distinguish the same items or similar items with basically the same function and effect.

[0039] The technical solution of the present application is applicable to the iris registration and iris recognition scenarios. The iris registration refers to the process of collecting the iris image of a user by a collection device, extracting, encoding and storing the features thereof to form a personal iris template. The iris recognition refers to the process of collecting the iris image of a person to be identified, extracting the features of the collected iris image and comparing the extracted features with the pre-registered iris feature template to verify the identity.

[0040] In the iris registration and iris recognition scenarios, the iris image is usually collected, the iris features in the iris image are extracted by a two-dimensional image processing algorithm (such as Gabor wavelet algorithm or local binary pattern (LBP) algorithm), the extracted iris features are compared with the pre-recorded iris feature template in the iris database to verify the user identity. The iris features extracted by the two-dimensional image processing algorithm usually reflect the two-dimensional texture features, and the criminals can counterfeit the iris image by the way of counterfeit sample to obtain the iris features, so as to pass the verification, and there is a risk of impersonation attack. In addition, since the pupil will shrink in the visible light environment, resulting in blurred texture, an infrared camera and a light source are needed for recognition, which increases the volume and cost of the recognition device; in strong light or relatively dark light, the accuracy of iris recognition is low.

[0041] To solve the above problems, the present application provides an iris registration scheme and an iris recognition scheme corresponding to the iris registration scheme. In the iris registration scheme, an iris image combination is formed by acquiring iris images of a user corresponding to different angles, three-dimensional iris point cloud data reflecting iris features of the user and spatial positions of the iris features are calculated based on the iris image combination, and a target three-dimensional iris model reflecting iris surface features of the user is generated by using the three-dimensional iris point cloud data. In the iris recognition scheme corresponding to the iris registration scheme, the current three-dimensional iris model of the user is verified by using the target three-dimensional iris model to complete identity verification. The iris surface features reflected by the target three-dimensional iris model not only include two-dimensional texture features, but also include depth information of the iris features, which increases the difficulty of iris feature forgery, helps to avoid impersonation attacks, and improves the security of the iris recognition technology. Since the current three-dimensional iris model of the user is verified by using the target three-dimensional iris model to complete verification, multiple images are acquired by using an ordinary camera to generate a three-dimensional iris model, and an infrared camera and a light source are not required for recognition, which helps to reduce the size and cost of the recognition device. In view of the problem that the accuracy of iris recognition under strong light or relatively dark light is low, the light parameters of the iris image acquisition device for acquiring the iris images are adjusted according to the environmental light intensity in the iris recognition process, which helps to improve the quality of the iris images and thus improves the accuracy of the iris recognition. In addition, the present application also provides a data fragmentation type storage scheme, in which an iris encryption template obtained by encrypting the target three-dimensional iris model and a key corresponding to the iris encryption template are segmented into multiple fragments and stored in different storage devices, so as to prevent the target three-dimensional iris model from being leaked and further improve the security of the iris recognition technology.

[0042] Among them, the iris registration scheme and the iris recognition scheme of the present application are respectively applied to an iris registration device and an iris recognition device. The iris registration device and the iris recognition device can be the same computer device, for example, the iris registration device and the iris recognition device are both cloud servers. The iris registration device and the iris recognition device can also be different computer devices, for example, the iris registration device is a mobile phone device or an augmented reality (AR) device, and the iris recognition device is a cloud server. The present application does not limit this.

[0043] The technical scheme of the present application will be specifically introduced below.

[0044] First, the iris registration scheme of the present application is introduced: Referring to Figure 1 , Figure 1 A flowchart of an iris registration method provided by an embodiment of the present application is shown in the figure. The method is applied to an iris registration device. As shown in Figure 1 , the method includes the following steps: S101, acquiring an iris image combination corresponding to a target user.

[0045] Here, the target user is understood as a user who needs to be authenticated, and the target user can be any user.

[0046] The target user's iris image combination includes multiple iris images of the target user, and the multiple iris images of the target user are iris images corresponding to different angles, which means the iris images of the target user under different angles of view.

[0047] For example, the iris image of the target user is collected through AR glasses. After the target user wears the AR glasses, the head posture of the target user is detected through the gyroscope in the AR glasses. During the process of acquiring the iris image, the target user is prompted to rotate the head within a range of 15° left and right, and adjust the eye angle of view within a range of 10° up and down. The angle of the user is detected each time the angle is adjusted. If the angle of the user is reasonable, the corresponding iris image is acquired. If the angle of the user is unreasonable, the user is prompted to adjust the angle again. The user adjusts the angle multiple times, and the corresponding iris image is acquired each time the angle is reasonable, so as to acquire the iris images corresponding to different angles. The reasonable angle can mean that the angle difference between the adjusted angle and the previous angle is less than a first angle, and the unreasonable angle can mean that the angle difference between the adjusted angle and the previous angle is greater than a second angle, and the second angle is greater than the first angle. For example, the first angle is 2°, and the second angle is 5°.

[0048] In some embodiments, multiple frames of iris images are acquired for each angle, and the target user's iris image combination is acquired through steps A1-A2 as follows: A1, acquire a plurality of iris image sets corresponding to the target user.

[0049] Here, the plurality of iris image sets are iris image sets corresponding to multiple angles, and each iris image set includes multiple frames of iris images of the target user under the same angle.

[0050] During the process of acquiring the iris image, when it is detected that the angle of the user is reasonable, multiple frames of images are acquired under the angle to form the iris image combination corresponding to the angle. The user is prompted to adjust the angle multiple times, and multiple frames of iris images are acquired each time the angle of the user is detected to be reasonable, so as to acquire the plurality of iris image sets corresponding to the target user.

[0051] For example, 3 frames of iris images are acquired each time the angle is reasonable, and the user is prompted to adjust the angle 5 times, so that 5 iris image sets can be obtained, and each iris image set in the 5 iris image sets includes 3 frames of iris images of the target user.

[0052] A2, image fusion is performed on the multiple frames of iris images in the same iris image set to obtain an iris image combination corresponding to the target user.

[0053] Here, the image fusion on the multiple frames of iris images in the same iris image set refers to that after image information (the image information includes but is not limited to color, texture) of each frame of iris images in the same iris image set is extracted, the extracted image information is comprehensively processed (such as taking the maximum value of pixels, weighted average value, etc.), and is merged into one frame of clearer iris image. The purpose of the image fusion on the multiple frames of iris images in the same iris image set is to make up for the limitations of a single frame of iris image and to improve the integrity and usability of image information.

[0054] After the image fusion on the multiple frames of iris images in one iris image set, an iris image corresponding to one angle is obtained. Since there are multiple iris image sets, multiple iris images corresponding to multiple angles are obtained, and the multiple iris images corresponding to the multiple angles constitute the iris image combination corresponding to the target user, that is, the iris image combination corresponding to the target user includes images obtained by respectively performing image fusion on the multiple iris image sets.

[0055] In a feasible implementation, the iris image combination includes an iris image corresponding to a target angle, and the target angle is one of the multiple angles; the image fusion is performed on the multiple frames of iris images in the same iris image set by the following steps a1-a3 to obtain the iris image combination corresponding to the target angle: a1, calculating an optical flow vector between two adjacent frames of iris images in the target iris image set.

[0056] Here, the target iris image set is an iris image set corresponding to the target angle.

[0057] For example, there are totally 5 iris image sets, and the 5 iris image sets correspond to 5 angles respectively, the target angle is any one of the 5 angles, and the target iris image set is any one of the 5 iris image sets.

[0058] The optical flow vector between the two adjacent frames of iris images includes optical flow vectors corresponding to each pixel point in the two adjacent frames of iris images, and the optical flow vector is used to reflect the movement amount of the pixel point in two directions of the image coordinate system.

[0059] The optical flow vector of one pixel point can be expressed as (Δu i , Δv i ), (Δu i , Δv i ) represents the optical flow vector of the i-th pixel point.

[0060] The optical flow vector between two adjacent frames of the target iris image set can be estimated by an optical flow estimation algorithm. The optical flow estimation algorithm includes but is not limited to Lucas-Kanade optical flow algorithm, Horn-Schuck optical flow algorithm, deep learning-based optical flow estimation algorithm, etc.

[0061] a2, registering each frame of the target iris image set based on the optical flow vector between two adjacent frames of the target iris image set.

[0062] Specifically, a reference iris image is determined in the target iris image set, and each frame of the target iris image set except the reference iris image is aligned to the coordinate system of the reference iris image according to the optical flow vector between two adjacent frames of the target iris image set, so as to register each frame of the target iris image set. The reference iris image can be any frame of the target iris image set.

[0063] For example, the target iris image set includes three consecutive frames of iris images, which are iris image P1, iris image P2 and iris image P3 in sequence, the time sequence of the iris image P1 is before that of the iris image P2, and the time sequence of the iris image P2 is before that of the iris image P3. Assuming that the iris image P2 is selected as the reference iris image, the optical flow vector between the iris image P2 and the iris image P1 and the optical flow vector between the iris image P3 and the iris image P2 are calculated by the foregoing step a1. For a pixel point in the iris image P1 (hereinafter referred to as a first pixel point), assuming that the pixel coordinates of the first pixel point in the iris image P1 are (u1, v1), the pixel coordinates of the first pixel point in the pixel coordinate system of the iris image P2 are (u2, v2), u2 = u1 + Δu1, v2 = v1 + Δv1, Δu1 and Δv1 represent the optical flow vector between the iris image P2 and the iris image P1, and the pixel points in the iris image P1 are converted to the coordinate system of the iris image P2, and the registration of the iris image P1 is completed. For a pixel point in the iris image P3 (hereinafter referred to as a second pixel point), assuming that the pixel coordinates of the second pixel point in the iris image P3 are (u3, v3), the pixel coordinates of the second pixel point in the pixel coordinate system of the iris image P2 are (u4, v4), u4 = u3 - Δu2, v4 = v3 - Δv2, Δu2 and Δv2 represent the optical flow vector between the iris image P3 and the iris image P2, and the pixel points in the iris image P3 are converted to the coordinate system of the iris image P2, and the registration of the iris image P3 is completed. In this way, all the pixel points of the iris images are converted to the same coordinate system, i.e. the registration of each frame of the target iris image set is completed, which facilitates subsequent operations in the same coordinate system.

[0064] a3, superimposed fusion is performed on each frame of the target iris image set after registration to obtain an iris image corresponding to the target angle.

[0065] wherein the mean value of the pixel values at the same pixel position in each frame of the target iris image set after registration can be calculated to obtain the iris image corresponding to the target angle.

[0066] The pixel value in the iris image corresponding to the target angle is represented as:

[0067] I d (u, v) represents the pixel value of the iris image corresponding to the target angle, I j (u, v) represents the pixel value of the jth frame of the target iris image set after registration, and n represents the number of iris images in the target iris image set.

[0068] In the above steps a1-a3, each frame of the iris image corresponding to the same angle is registered based on the optical flow vector between the adjacent two frames of the iris image, and the superimposed fusion is performed on the iris image after registration, which can ensure that the homologous pixels (such as the same object point) are accurately superimposed during fusion, reduce artifacts, and improve the image quality of the fused iris image.

[0069] For each of the plurality of iris image sets, the image fusion can be performed on the plurality of frames of the iris image set in the manner introduced in the above steps A1-A2, so as to obtain the iris image corresponding to each angle.

[0070] Alternatively, the image fusion can also be performed on the plurality of frames of the iris image in the same iris image set based on the time domain average algorithm to obtain an iris image combination corresponding to one angle. The present application does not limit this.

[0071] In the above steps A1-A2, for each angle, the plurality of frames of the iris image are acquired for image fusion to obtain the iris image corresponding to each angle to form the iris image combination, which can improve the quality of the iris image and obtain the iris image with higher resolution.

[0072] In some embodiments, before the image fusion is performed on the plurality of frames of the iris image in the same iris image set to obtain the iris image combination corresponding to the target user, the quality of the iris image in the target iris image set can be detected, and in the case that there is an iris image in the target iris image set that does not meet the preset quality condition, the iris image of the target user under the target angle is re-acquired until all the iris images in the target iris image set meet the preset quality condition.

[0073] wherein the preset quality condition includes at least one or more of the following conditions: (1) the Laplace transform variance of the iris image is less than a variance threshold, for example, 200; (2) the ratio of the pupil diameter in the iris image to the image width of the iris image is greater than a ratio threshold, for example, 12%; (3) the luminance difference ratios between the four corner regions and the central region of the iris image are all less than a difference threshold, the corner regions and the central region contain k*k pixels, the calculation formula of the luminance difference ratio is: luminance difference ratio = (luminance average value of the corner region - luminance average value of the central region) / luminance average value of the central region * 100%, for example, the difference threshold is 15%.

[0074] After obtaining the iris image of the target user at the target angle, the Laplace transform of the iris image of the target user at the target angle is performed, and then the pixel variance of the transformed image is calculated, so as to obtain the Laplace transform variance of the iris image of the target user at the target angle.

[0075] After obtaining the iris image of the target user at the target angle, the pupil region in the iris image of the target user at the target angle is located based on the pupil positioning technology, and then the center and radius of the pupil are determined through the Hough circle detection, so as to calculate the pupil diameter, and then calculate the ratio of the pupil diameter to the image width, so as to obtain the ratio of the pupil diameter to the image width in the iris image of the target user at the target angle.

[0076] After obtaining the iris image of the target user at the target angle, the four corner regions and the central region in the iris image of the target user at the target angle are determined, and then the luminance average values of the four corner regions and the central region are calculated, and then the luminance difference ratios between the four corner regions and the central region are calculated based on the above calculation formula of the luminance difference ratio, so as to obtain the luminance difference ratios between the four corner regions and the central region of the iris image of the target user at the target angle.

[0077] Then, according to the Laplace transform variance of the iris image of the target user at the target angle, and / or, the ratio of the pupil diameter to the image width in the iris image of the target user at the target angle, and / or, the luminance difference ratios between the four corner regions and the central region of the iris image of the target user at the target angle, it is judged whether the iris image of the target user at the target angle satisfies the preset quality condition.

[0078] In the case that the iris image does not satisfy the preset quality condition, the iris image is reacquired, which can ensure the quality of the iris image.

[0079] S102, based on the iris image corresponding to the target user, three-dimensional iris point cloud data is calculated.

[0080] Here, the three-dimensional iris point cloud data is used to reflect the iris features of the target user and the spatial positions of the iris features of the target user. The three-dimensional iris point cloud data includes spatial position coordinates of the iris feature points and feature values of the iris feature points, and the feature values of the iris feature points are obtained by performing iris feature extraction on the iris images.

[0081] In some implementations, the three-dimensional iris point cloud data is calculated through steps B1-B3 as follows: B1, performing iris feature extraction on each iris image in the iris image combination corresponding to the target user to obtain iris feature points corresponding to the plurality of iris images in the iris image combination.

[0082] The iris feature extraction algorithm includes but is not limited to a scale-invariant feature transform (SIFT) algorithm, a speeded up robust features (SURF) algorithm, an oriented FAST and rotated BRIEF (ORB) algorithm, and the like.

[0083] The iris feature extraction algorithm is used to perform iris feature extraction on each iris image in the iris image combination corresponding to the target user, so that the position of the iris feature point in each iris image and the feature descriptor corresponding to the iris feature point can be extracted. The feature descriptor is an attribute parameter used to describe the iris feature point, and is usually a multidimensional position vector.

[0084] B2, performing feature matching on the iris feature points corresponding to the plurality of iris images in the iris image combination to obtain a plurality of groups of iris feature matching pairs.

[0085] Here, each group of iris feature matching pairs includes two matched iris feature points. The two matched iris feature points in the iris feature matching pair are iris feature points corresponding to each other in two iris images in the iris image combination, and the two iris feature points in the iris feature matching pair correspond to the same object point (i.e., a point on the iris).

[0086] In a specific implementation, for two of the iris image combination (hereinafter referred to as the first iris image and the second iris image), the feature descriptors of the iris feature points (hereinafter referred to as the first iris feature points) in the first iris image can be traversed and obtained; the Euclidean distances between the feature descriptors of the first iris feature points and the feature descriptors of the iris feature points in the second iris image are calculated, if the Euclidean distance between the feature descriptor of the first iris feature point and the feature descriptor of the second iris feature point in the second iris image is less than the preset distance threshold, it is determined that the first iris feature point and the second iris feature point are matched, and the first iris feature point and the second iris feature point are determined as a group of iris feature matching pairs; if the Euclidean distance between the feature descriptor of the first iris feature point and the feature descriptor of each iris feature point in the second iris image is greater than or equal to the preset distance threshold, the next iris feature point in the first iris image is traversed and feature matching is performed. In this way, the iris feature matching pairs corresponding to the two iris images in the iris image combination are found.

[0087] For a plurality of iris images in the iris image combination, the iris images are combined two by two, and the iris feature matching pairs are determined in the manner of the foregoing feature matching, so as to obtain a plurality of groups of iris feature matching pairs.

[0088] B3, based on the plurality of groups of iris feature point pairs, three-dimensional iris point cloud data is calculated.

[0089] Wherein, the spatial position coordinates of the iris feature points and the iris feature point pairs satisfy the following relationship: P = λ1*x1 = λ2*(RP+t), , p1 = (u1, v1), p2 = (u2, v2).

[0090] P represents the spatial position coordinates of the iris feature points, (u1, v1) and (u2, v2) represent the position coordinates of the two iris feature points in the two iris images, x1 and x2 represent the corresponding homogeneous coordinates of the two iris feature points in the iris feature point pair, K represents the intrinsic parameter of the iris image acquisition device for obtaining the iris image, p1 and p2 represent the two iris feature points in the iris feature point pair, and R and t represent the extrinsic parameter of the iris image acquisition device for obtaining the iris image.

[0091] The spatial position coordinates of the iris feature points can be obtained by solving λ1, λ2 and P by the least square method.

[0092] The spatial position coordinates of the plurality of iris feature points are determined in the same manner, and the spatial position coordinates of the plurality of iris feature points are obtained; the feature descriptors of the iris feature points are taken as the feature values of the iris feature points, and the feature values of the plurality of iris feature points are obtained; the spatial position coordinates of the plurality of iris feature points and the feature values of the plurality of iris feature points form the three-dimensional iris point cloud data.

[0093] In the above steps B1-B3, the three-dimensional iris point cloud data is calculated based on the iris feature matching manner, which is less affected by light and helps to accurately determine the spatial position of the iris feature.

[0094] S103, generating a target three-dimensional iris model of the target user according to the three-dimensional iris point cloud data.

[0095] Here, the target three-dimensional iris model of the target user is used to reflect the iris surface feature of the target user, and the iris surface feature includes the depth information of the iris feature points in addition to the feature values of the iris feature points (the feature values are used to reflect the plane feature).

[0096] In some embodiments, the three-dimensional iris point cloud data is surface reconstructed to obtain the target three-dimensional iris model of the target user. Wherein, the three-dimensional iris point cloud data can be surface reconstructed based on a triangulation or curve fitting algorithm to obtain the target three-dimensional iris model of the target user.

[0097] After the three-dimensional iris point cloud data is surface reconstructed, the feature values of each iris feature point in the three-dimensional iris point cloud data are saved; and the pixel values and spatial position coordinates corresponding to each iris feature point in the three-dimensional iris point cloud data are saved, to obtain the target three-dimensional iris model of the target user. The target three-dimensional iris model is the personal iris template of the target user.

[0098] The target three-dimensional iris model is obtained through surface reconstruction, so that the target three-dimensional iris model can reflect the iris surface feature of the user, thereby covering the depth information of the iris feature.

[0099] In the above Figure 1In the corresponding technical solution, a target user's iris image combination is obtained, the iris image combination including iris images corresponding to different angles of the target user, then based on the iris image combination, three-dimensional iris point cloud data is calculated, the three-dimensional iris point cloud data being used to reflect iris features and spatial positions of the iris features of the target user, and finally, a target three-dimensional iris model of the target user is generated according to the three-dimensional iris point cloud data, the target three-dimensional iris model being used to reflect iris surface features of the target user. Since the target three-dimensional iris model of the user is generated in advance, the target three-dimensional iris model reflects the iris surface features of the user, and compared with two-dimensional iris features, the iris surface features further include depth information of the iris features. Therefore, when the iris features obtained by means of a fake sample (such as a photo or contact lens projection) are compared with the target three-dimensional iris model, the iris features cannot pass the verification due to the lack of the depth information of the iris features, thus the security of the iris recognition technology is improved.

[0100] After the target three-dimensional iris model of the target user is generated, the target three-dimensional iris model of the target user can be stored in a preset iris database, the iris database being used to store target three-dimensional iris models of different users, and the target three-dimensional iris models stored in the iris database being used for iris recognition. The target three-dimensional iris model of the target user can be saved in the preset iris database in plaintext or in ciphertext. The target three-dimensional iris model of the target user can also be stored in multiple preset storage devices in the form of slices, the multiple preset storage devices being storage devices preset for distributed storage of iris encryption templates. For the implementation of storing the target three-dimensional iris model of the target user in the multiple preset storage devices in the form of slices, please refer to the following Figure 2 The corresponding description is provided. In a specific implementation scenario, the multiple preset storage devices include a mobile phone device of the target user, AR glasses of the target user, and a cloud server.

[0101] Referring to Figure 2 , Figure 2 Another iris registration method provided by the embodiments of the present application is provided, and the method is applied to an iris registration device. As shown in Figure 2 , the method includes the following steps: S201, obtaining an iris image combination corresponding to a target user.

[0102] S202, calculating three-dimensional iris point cloud data based on the iris image combination corresponding to the target user.

[0103] S203, generating a target three-dimensional iris model of the target user according to the three-dimensional iris point cloud data.

[0104] Here, the specific implementation of steps S201-S203 can refer to the description of the aforementioned steps S101-S103, and will not be described here again.

[0105] S204, encrypting the target three-dimensional iris model of the target user to obtain an iris encryption template corresponding to the target user.

[0106] The target three-dimensional iris model of the target user can be encrypted based on a symmetric encryption algorithm (such as the SM4 algorithm) or an asymmetric encryption algorithm (such as the SM2 algorithm) to obtain the iris encryption template corresponding to the target user.

[0107] S205, segmenting the iris encryption template corresponding to the target user to obtain a plurality of iris encryption template fragments corresponding to the target user.

[0108] The iris encryption template corresponding to the target user can be segmented into m fragments to obtain a plurality of iris encryption template fragments corresponding to the target user, and m is the number of the plurality of preset storage devices.

[0109] Specifically, the iris encryption template corresponding to the target user can be segmented into m fragments in an equal division manner, and each fragment contains a data amount of M / m, and M is the total data amount contained in the iris encryption template corresponding to the target user. For example, the length of the iris encryption template corresponding to the target user is 3MB, and there are 3 preset storage devices, so the iris encryption template corresponding to the target user can be equally divided into 3 fragments of 1MB.

[0110] Alternatively, the iris encryption template corresponding to the target user can also be segmented into m fragments based on the Shamir secret sharing algorithm, which is not limited in the present application.

[0111] S206, saving different iris encryption template fragments in the plurality of iris encryption template fragments corresponding to the target user to different plurality of preset storage devices.

[0112] One preset storage device is used to store one iris encryption template fragment, and different preset storage devices are used to store different iris encryption template fragments.

[0113] After the plurality of iris encryption template fragments corresponding to the target user are segmented, the user identifier of the target user can be used as a fragment identifier of the plurality of iris encryption template fragments corresponding to the target user to identify the iris encryption template fragments corresponding to the target user, the plurality of iris encryption template fragments corresponding to the target user are distributed to the plurality of preset storage devices, and the user identifier of the target user is sent to the plurality of preset storage devices; the plurality of preset storage devices save the user identifier of the target user and the iris encryption template fragments corresponding to the target user received by each of the plurality of preset storage devices.

[0114] After distributing the plurality of iris encryption template fragments corresponding to the target user to the plurality of preset storage devices and sending the user identifier of the target user to the plurality of preset storage devices, the user identifier of the target user, algorithm related information for encrypting the iris encryption template corresponding to the target user, and storage locations of the plurality of iris encryption template fragments can be correspondingly stored in a storage mapping table. The algorithm related information includes algorithm name, algorithm version, and the like.

[0115] In the above steps S204-S206, after the target three-dimensional iris model is encrypted and segmented into a plurality of iris encryption template fragments and then distributed for storage, the target three-dimensional iris model is prevented from being leaked, and the security of the iris recognition technology is improved.

[0116] It can be understood that after the target three-dimensional iris model of the target user is encrypted to obtain the iris encryption template corresponding to the target user, the iris encryption template corresponding to the target user can also be directly stored in a preset iris database.

[0117] In some embodiments, after the target three-dimensional iris model of the target user is encrypted to obtain the iris encryption template corresponding to the target user, the following steps S207-S208 are further included: S207, the key corresponding to the iris encryption template corresponding to the target user is segmented to obtain a plurality of key fragments corresponding to the target user.

[0118] Here, if the target three-dimensional iris model of the target user is encrypted using a symmetric encryption algorithm to obtain the iris encryption template corresponding to the target user, the key corresponding to the iris encryption template corresponding to the target user is a symmetric key, that is, the key for encrypting the iris encryption template corresponding to the target user; if the target three-dimensional iris model of the target user is encrypted using an asymmetric algorithm to obtain the iris encryption template corresponding to the target user, the key corresponding to the iris encryption template corresponding to the target user is a private key in a non-symmetric key pair, and the key for encrypting the iris encryption template corresponding to the target user is a public key in the non-symmetric key pair.

[0119] Among them, the key corresponding to the iris encryption template corresponding to the target user can be segmented into m fragments to obtain a plurality of key fragments corresponding to the target user.

[0120] S208, different key fragments in the plurality of key fragments corresponding to the target user are stored in different plurality of preset storage devices.

[0121] Among them, one preset storage device is used to store one key fragment, and different preset storage devices are used to store different iris encryption template fragments and different key fragments.

[0122] After the target user's corresponding multiple key fragments are obtained by segmentation, the user identifier of the target user can be used as the fragment identifier of the target user's corresponding multiple iris encryption template fragments to identify the target user's corresponding key fragments, the target user's corresponding multiple key fragments are distributed to the multiple preset storage devices, and the user identifier of the target user is sent to the multiple preset storage devices; the multiple preset storage devices save the user identifier of the target user and the target user's corresponding key fragments received by each of the multiple preset storage devices.

[0123] In the above steps S207-S208, after the key corresponding to the iris encryption template is segmented into multiple key fragments and distributed for storage, the decryption difficulty of the iris encryption template is increased, and the target three-dimensional iris model is further prevented from being leaked.

[0124] It can be understood that after the target three-dimensional iris model of the target user is encrypted to obtain the target user's corresponding iris encryption template, the key corresponding to the target user's corresponding iris encryption template can also be stored locally.

[0125] Next, the iris recognition scheme of the present application is introduced: Referring to Figure 3 , Figure 3 A flowchart of an iris recognition method provided by an embodiment of the present application is shown in the figure. The method is applied to an iris recognition device; as shown in Figure 3 the figure, the method includes the following steps: S301, obtaining a current three-dimensional iris model of a target user.

[0126] Here, the definition of the target user is described with reference to the foregoing step S101.

[0127] The current three-dimensional iris model of the target user is used to reflect the iris surface features of the target user. The iris surface features are described with reference to the foregoing step S103.

[0128] In some embodiments, the current three-dimensional iris model of the target user is obtained by the following steps C1-C3: C1, obtaining multiple iris images of the target user.

[0129] Among them, the multiple iris images of the target user can be obtained by an iris image acquisition device; the iris image acquisition device is an acquisition device for acquiring iris images for iris recognition, and the specific type of the iris image acquisition device depends on the specific application scenario of iris recognition. The iris image acquisition device is, for example, AR glasses or a mobile phone device, etc.

[0130] In some embodiments, before obtaining the multiple iris images of the target user, the light parameter of the iris image acquisition device can also be adjusted according to the ambient light intensity.

[0131] wherein the light compensation parameter of the iris image acquisition device is adjusted when it is detected that the ambient light brightness is less than a first brightness value or greater than a second brightness value. The second brightness value is greater than the first brightness value, for example, the first brightness value is 50 lux, and the second brightness value is 1000 lux.

[0132] Specifically, when it is detected that the ambient light brightness is less than a first brightness value or greater than a second brightness value, the light compensation parameter of the iris image acquisition device can be adjusted according to a preset light compensation parameter mapping model. The light compensation parameter mapping model can be a linear mapping model or a piecewise function model, and the present application does not limit this.

[0133] According to the ambient light intensity, the light parameter of the iris image acquisition device is adjusted, which helps to improve the quality of the iris image.

[0134] In some embodiments, before acquiring the plurality of iris images of the target user, the target user can also be subjected to a living body detection. When the living body detection passes, the plurality of iris images of the target user are acquired.

[0135] C2, based on the plurality of iris images of the target user, current three-dimensional iris point cloud data is calculated.

[0136] Here, the current three-dimensional iris point cloud data is used to reflect the iris features of the target user and the spatial position of the iris features of the target user.

[0137] Regarding the calculation of the current three-dimensional iris point cloud data based on the plurality of iris images of the target user, the same as the calculation of the three-dimensional iris point cloud data based on the combination of the corresponding iris images of the target user in the foregoing step S102, please refer to the description of the foregoing step S102, which will not be repeated here.

[0138] C3, according to the current three-dimensional iris point cloud data, a current three-dimensional iris model of the target user is generated.

[0139] Here, according to the current three-dimensional iris point cloud data, a current three-dimensional iris model of the target user is generated, which is the same as the generation of the target three-dimensional iris model of the target user according to the three-dimensional iris point cloud data in the foregoing step S103. Please refer to the description of the foregoing step S103, which will not be repeated here.

[0140] S302, a target three-dimensional iris model of the target user is acquired.

[0141] Here, the target three-dimensional iris model of the target user is generated in advance by the foregoing Figure 1 or Figure 2 iris registration scheme introduced.

[0142] In some embodiments, in the case that the target three-dimensional iris model of the target user is stored in the preset iris database in plaintext, the target three-dimensional iris model corresponding to the user identifier of the target user can be obtained from the preset iris database as the target three-dimensional iris model of the target user.

[0143] In other embodiments, in the case that the target three-dimensional iris model of the target user is stored in the manner introduced in the foregoing steps S204-S206, the target three-dimensional iris model of the target user can be obtained through the following steps D1-D3: D1, obtain the iris encryption template fragments corresponding to the target user from different preset storage devices to obtain a plurality of iris encryption template fragments corresponding to the target user.

[0144] Here, the definition of the plurality of preset storage devices can refer to the foregoing description.

[0145] Among them, the plurality of preset storage devices can be determined according to the storage location of the user identifier of the target user in the storage mapping table, the user identifier of the target user is sent to each preset storage device of the plurality of preset storage devices, so that the plurality of preset storage devices find the iris encryption template fragments corresponding to the target user according to the user identifier of the target user, and the iris encryption template fragments corresponding to the target user sent by the plurality of preset storage devices are received to obtain the plurality of iris encryption template fragments corresponding to the target user.

[0146] D2, combine the plurality of iris encryption template fragments corresponding to the target user to obtain the iris encryption template corresponding to the target user.

[0147] Among them, the plurality of iris encryption template fragments corresponding to the target user can be spliced to obtain the iris encryption template corresponding to the target user.

[0148] D3, decrypt the iris encryption template corresponding to the target user to obtain the target three-dimensional iris model of the target user.

[0149] In some embodiments, in the case that the key corresponding to the iris encryption template corresponding to the target user is stored locally, the key corresponding to the iris encryption template corresponding to the target user can be obtained from the local, and the key obtained is used to decrypt the iris encryption template corresponding to the target user based on the algorithm-related information in the storage mapping table to obtain the target three-dimensional iris model of the target user.

[0150] In other embodiments, in the case that the key corresponding to the iris encryption template corresponding to the target user is stored in the manner introduced in the foregoing steps S207-S208, the iris encryption template is decrypted through the following steps E1-E3 to obtain the target three-dimensional iris model of the target user: E1, acquire the key fragments corresponding to the target user from different preset storage devices, to obtain a plurality of key fragments corresponding to the target user.

[0151] Wherein, the plurality of preset storage devices can be determined according to the storage location of the user identifier of the target user in the storage mapping table, the user identifier of the target user is sent to each preset storage device of the plurality of preset storage devices, so that the plurality of preset storage devices find the key fragments corresponding to the target user according to the user identifier of the target user, and receive the key fragments corresponding to the target user sent by the plurality of preset storage devices, so as to obtain the plurality of key fragments corresponding to the target user.

[0152] E2, combine the plurality of key fragments corresponding to the target user, to obtain the key corresponding to the iris encryption template of the target user.

[0153] Wherein, the plurality of key fragments corresponding to the target user can be spliced to obtain the key corresponding to the iris encryption template of the target user.

[0154] E3, decrypt the iris encryption template of the target user using the key corresponding to the iris encryption template of the target user, to obtain the target three-dimensional iris model of the target user.

[0155] Wherein, based on the algorithm related information in the storage mapping table, the iris encryption template of the target user is decrypted using the key corresponding to the iris encryption template of the target user, to obtain the target three-dimensional iris model of the target user.

[0156] S303, compare the current three-dimensional iris model of the target user with the target three-dimensional iris model of the target user, to obtain the iris recognition result of the target user.

[0157] Wherein, the similarity between the current three-dimensional iris model of the target user and the target three-dimensional iris model of the target user can be calculated; if the similarity is greater than a similarity threshold, it is determined that the iris recognition result of the target user is verified; if the similarity is less than or equal to the similarity threshold, it is determined that the iris recognition result of the target user is not verified.

[0158] In the above Figure 3Corresponding technical solutions, after obtaining the current three-dimensional iris model of the target user, the target three-dimensional iris model of the target user is obtained, and the current three-dimensional iris model and the target three-dimensional iris model are compared to obtain the iris recognition result of the target user; Since the target three-dimensional iris model of the user is generated in advance, the target three-dimensional iris model reflects the iris surface characteristics of the user, and compared with the two-dimensional iris characteristics, the depth information of the iris characteristics is also covered in the iris surface characteristics. The iris characteristics obtained by the counterfeit sample (such as photo, contact lens projection, etc.) cannot be verified when compared with the target three-dimensional iris model due to the lack of depth information of the iris characteristics, so that the security of the iris recognition technology can be improved.

[0159] The method of the present application is introduced above, and the device of the present application is introduced below.

[0160] Referring to Figure 4 , Figure 4 is a structural schematic diagram of an iris registration device provided by an embodiment of the present application, as shown in Figure 4 , the iris registration device 40 comprises: an iris image acquisition module 401, configured to acquire a target user corresponding iris image combination, the iris image combination comprising a plurality of iris images of the target user, the plurality of iris images being iris images corresponding to different angles; an iris point cloud calculation module 402, configured to calculate three-dimensional iris point cloud data based on the iris image combination, the three-dimensional iris point cloud data being used to reflect iris characteristics of the target user and a spatial position of the iris characteristics; an iris template generation module 403, configured to generate a target three-dimensional iris model of the target user according to the three-dimensional iris point cloud data, the target three-dimensional iris model being used to reflect iris surface characteristics of the target user.

[0161] It should be noted that the above-mentioned iris registration device 40 can execute the above-mentioned iris registration method provided by the present application, and has the corresponding function modules and beneficial effects of the execution method. Technical details not described in detail in the embodiments can be referred to the iris registration method provided by the present application.

[0162] Referring to Figure 5 , Figure 5 is a structural schematic diagram of an iris recognition device provided by an embodiment of the present application, as shown in Figure 5 , the iris recognition device 50 comprises: an iris model acquisition module 501, configured to acquire a current three-dimensional iris model of a target user, the current three-dimensional iris model being used to reflect iris surface characteristics of the target user; The iris template acquisition module 502 is configured to acquire a target three-dimensional iris model of the target user, the target three-dimensional iris model of the target user being generated in advance by the iris registration method described above. The iris recognition module 503 is configured to compare the current three-dimensional iris model with the target three-dimensional iris model to obtain an iris recognition result of the target user.

[0163] It should be noted that the iris recognition apparatus 50 can perform the iris recognition method provided in the embodiments of the present application, and has the function modules and beneficial effects corresponding to the execution method. Technical details not described in the embodiments can be referred to the iris recognition method provided in the embodiments of the present application.

[0164] Referring to Figure 6 , Figure 6 FIG. 6 is a structural schematic diagram of a computer device provided in an embodiment of the present application. The computer device 60 includes a processor 601 and a memory 602. The memory 602 is connected to the processor 601, for example, by a bus.

[0165] The processor 601 is configured to support the computer device 60 to perform the functions corresponding to the method in the method embodiments described above. The processor 601 can be a central processing unit (CPU), a network processor (NP), a hardware chip or any combination thereof. The hardware chip can be an application specific integrated circuit (ASIC), a programmable logic device (PLD) or a combination thereof. The PLD can be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL) or any combination thereof.

[0166] The memory 602 is configured to store program codes and the like. The memory 602 can include a volatile memory (VM), such as a random access memory (RAM), and / or a non-volatile memory (NVM), such as a read-only memory (ROM), a flash memory, a hard disk drive (HDD), or a solid-state drive (SSD), and / or a combination thereof.

[0167] The memory 602 is configured to store non-volatile software programs, non-volatile computer-executable programs and modules, such as program instructions / modules corresponding to the iris registration method or the iris recognition method in the embodiments of the present application. The processor executes various functional applications and data processing of the iris registration method or the iris recognition method by running the non-volatile software programs, instructions and modules stored in the memory, that is, realizes the functions of the iris registration method or the iris recognition method provided by the method embodiments.

[0168] The memory 602 can include a program storage area and a data storage area. The program storage area can store an operating system and application programs required by at least one function. The data storage area can store data created according to the use of the iris registration device or the iris recognition device, and the like. In some embodiments, the memory can include a memory remotely arranged with respect to the processor, and these remote memories can be connected to the iris registration device or the iris recognition device through a network. Examples of the network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0169] The one or more modules are stored in the memory and, when executed by the one or more processors, perform the iris registration method or the iris recognition method in any of the above method embodiments, for example, perform the method steps described in the above method embodiments, and realize the functions of the modules described in the above device embodiments.

[0170] The embodiments of the present application also provide a computer readable storage medium storing a computer program, the computer program including program instructions, the program instructions causing a computer to perform the iris registration method or the iris recognition method as described in the foregoing embodiments when the computer program is executed by the computer.

[0171] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the associated hardware through a computer program. The program can be stored in a computer readable storage medium, and when the program is executed, the processes of the above-mentioned embodiments of the methods can be included. The storage medium can be a magnetic disc, an optical disc, a read-only memory (ROM), a random access memory (RAM), or the like.

[0172] The above disclosure is only the preferred embodiment of the present application, and of course cannot limit the scope of the right of the present application, so the equivalent changes made according to the claims of the present application still belong to the scope covered by the present application.

Claims

1. An iris registration method, characterized in that, include: Obtain a combination of iris images corresponding to a target user, wherein the combination of iris images includes multiple iris images of the target user, and the multiple iris images are iris images corresponding to different angles; Based on the iris image combination, three-dimensional iris point cloud data is calculated. The three-dimensional iris point cloud data is used to reflect the iris features of the target user and the spatial location of the iris features. Based on the three-dimensional iris point cloud data, a target three-dimensional iris model of the target user is generated, which is used to reflect the iris surface features of the target user.

2. The method according to claim 1, characterized in that, The acquisition of the iris image combination corresponding to the target user includes: Obtain multiple iris image sets corresponding to the target user. The multiple iris image sets are iris image sets corresponding to multiple angles. Each iris image set includes multiple frames of iris images of the target user at the same angle. Multiple iris images from the same iris image set are fused to obtain the iris image combination. The multiple iris images in the iris image combination include images obtained by fusing the multiple iris image sets separately.

3. The method according to claim 2, characterized in that, The iris image combination includes an iris image corresponding to a target angle, where the target angle is one of the plurality of angles; The process of fusing multiple iris images from the same iris image set to obtain the iris image combination includes: Calculate the optical flow vector between two adjacent iris images in the target iris image set, wherein the target iris image set is the iris image set corresponding to the target angle; Based on the optical flow vector, each frame of the iris image in the target iris image set is registered; The registered iris images are superimposed and fused to obtain the iris image corresponding to the target angle.

4. The method according to claim 3, characterized in that, Before performing image fusion on multiple iris images from the same iris image set, the method further includes: If there are iris images in the target iris image set that do not meet the preset quality conditions, the iris image of the target user at the target angle is reacquired until all iris images in the target iris image set meet the preset quality conditions.

5. The method according to claim 1, characterized in that, The calculation of three-dimensional iris point cloud data based on the iris image combination includes: Iris features are extracted from each iris image in the iris image combination to obtain iris feature points corresponding to the multiple iris images; Feature matching is performed on the iris feature points corresponding to the multiple iris images to obtain multiple sets of iris feature matching pairs. Each set of iris feature matching pairs includes two matching iris feature points, which are the corresponding iris feature points in the two iris images in the iris image combination. Based on the multiple sets of iris feature point pairs, three-dimensional iris point cloud data is calculated.

6. The method according to claim 1, characterized in that, After generating the target 3D iris model of the target user based on the 3D iris point cloud data, the method further includes: The target three-dimensional iris model is encrypted to obtain the iris encryption template corresponding to the target user; The iris encryption template is segmented to obtain multiple iris encryption template fragments corresponding to the target user; The different iris encryption template segments in the multiple iris encryption template segments are saved to different preset storage devices.

7. The method according to claim 6, characterized in that, After encrypting the target 3D iris model to obtain the iris encryption template corresponding to the target user, the method further includes: The key corresponding to the iris encryption template is segmented to obtain multiple key fragments corresponding to the target user; The different key fragments from the multiple key fragments are saved to the multiple different preset storage devices.

8. An iris recognition method, characterized in that, include: Obtain the current three-dimensional iris model of the target user, wherein the current three-dimensional iris model is used to reflect the iris surface features of the target user; Obtain the target three-dimensional iris model of the target user, wherein the target three-dimensional iris model of the target user is pre-generated by the method described in any one of claims 1-7; The current 3D iris model is compared with the target 3D iris model to obtain the iris recognition result of the target user.

9. The method according to claim 8, characterized in that, The process of obtaining the target three-dimensional iris model of the target user includes: Iris encryption template fragments corresponding to the target user are obtained from multiple different preset storage devices to obtain multiple iris encryption template fragments corresponding to the target user; By combining the multiple iris encryption template fragments, the iris encryption template corresponding to the target user is obtained; The iris encryption template is decrypted to obtain the target three-dimensional iris model of the target user.

10. The method according to claim 9, characterized in that, The step of decrypting the iris encryption template to obtain the target three-dimensional iris model of the target user includes: Obtain the key fragments corresponding to the target user from the different preset storage devices to obtain multiple key fragments corresponding to the target user; By combining the multiple key fragments, the key corresponding to the iris encryption template is obtained; The iris encryption template is decrypted using the key to obtain the target three-dimensional iris model of the target user.

11. A computer device, characterized in that, The device includes a memory and a processor, the memory being connected to the processor, the processor being configured to execute one or more computer programs stored in the memory, the processor causing the computer device to perform the method as described in any one of claims 1-7 or any one of claims 8-10 when executing the one or more computer programs.

12. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, the computer program including program instructions that, when executed by a processor, cause the processor to perform the method as described in any one of claims 1-7 or any one of claims 8-10.

Images