Iris recognition method, device, apparatus and system

Abstract

The application provides an iris recognition method, device, equipment and system, the method comprising: acquiring at least two frames of to-be-recognized images under different light supplement states; using a deep learning algorithm to perform feature extraction and feature fusion processing on the at least two frames of to-be-recognized images to obtain fused features; and performing iris recognition according to the fused features. The method can improve the performance of iris recognition.

Description

Technical Field

[0001] This application relates to the field of biometric identification, and in particular to an iris recognition method, device, equipment and system. Background Technology

[0002] Iris recognition technology uses information from the iris region of the eye for identity authentication and is now widely used in many fields such as access control, security checks, security monitoring, and financial payments. Because iris recognition often requires infrared illumination to capture a clear iris image, in unconstrained scenarios, reflective objects such as glasses worn by the person being recognized may cause light spots in the iris image captured by the camera, thus affecting the performance of the entire iris recognition system. Summary of the Invention

[0003] In view of this, this application provides an iris recognition method, apparatus, device and system.

[0004] Specifically, this application is implemented through the following technical solution:

[0005] According to a first aspect of the embodiments of this application, an iris recognition method is provided, comprising:

[0006] Acquire at least two frames of images to be identified under different supplementary lighting conditions;

[0007] Using deep learning algorithms, feature extraction and feature fusion processing are performed on the at least two frames of images to be identified to obtain fused features;

[0008] Iris recognition is performed based on the fusion features.

[0009] According to a second aspect of the embodiments of this application, an iris recognition device is provided, comprising:

[0010] The acquisition unit is used to acquire at least two frames of images to be identified under different supplementary lighting conditions;

[0011] The feature processing unit is used to perform feature extraction and feature fusion processing on the at least two frames of images to be identified using deep learning algorithms to obtain fused features;

[0012] The identification unit is used to perform iris recognition based on the fusion features.

[0013] According to a third aspect of the present application, an electronic device is provided, including a processor and a memory, the memory storing machine-executable instructions executable by the processor, the processor being configured to execute the machine-executable instructions to implement the method provided in the first aspect.

[0014] According to a fourth aspect of the embodiments of this application, a machine-readable storage medium is provided, wherein machine-executable instructions are stored therein, and when the machine-executable instructions are executed by a processor, the method provided in the first aspect is implemented.

[0015] According to a fifth aspect of the embodiments of this application, an iris recognition system is provided, comprising: an image acquisition device, a supplementary lighting device, and an iris recognition device; wherein:

[0016] The image acquisition device is used for image acquisition;

[0017] The supplemental lighting device is used for providing supplemental lighting;

[0018] The iris recognition device is used to perform iris recognition according to the method provided in the first aspect.

[0019] The technical solution provided in this application can bring at least the following beneficial effects:

[0020] By acquiring at least two frames of images to be identified under different lighting conditions, and using deep learning algorithms to extract and fuse features from these at least two frames, information complementarity between images under different lighting conditions is achieved through high-level semantic fusion of the feature domain. While achieving information complementarity, the increase in noise caused by image fusion is avoided. Furthermore, by performing iris recognition based on the fused features, the performance of iris recognition is improved. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating an iris recognition method according to an exemplary embodiment of this application;

[0022] Figure 2 This is a schematic diagram illustrating a process for obtaining fusion features according to an exemplary embodiment of this application;

[0023] Figure 3 This is a schematic diagram illustrating the deployment of an iris recognition scenario in an exemplary embodiment of this application;

[0024] Figure 4 This is a schematic diagram illustrating an iris recognition implementation process according to an exemplary embodiment of this application;

[0025] Figure 5 This is a schematic diagram illustrating an iris recognition implementation process according to an exemplary embodiment of this application;

[0026] Figure 6 This is a schematic diagram of the structure of an iris recognition device shown in an exemplary embodiment of this application;

[0027] Figure 7This is a schematic diagram of the hardware structure of an electronic device illustrated in an exemplary embodiment of this application;

[0028] Figure 8 This is a schematic diagram of the architecture of an iris recognition system shown in an exemplary embodiment of this application. Detailed Implementation

[0029] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0030] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise.

[0031] To enable those skilled in the art to better understand the technical solutions provided in the embodiments of this application, and to make the above-mentioned objectives, features and advantages of the embodiments of this application more apparent and understandable, the technical solutions in the embodiments of this application will be further described in detail below with reference to the accompanying drawings.

[0032] Please see Figure 1 This is a flowchart illustrating an iris recognition method provided in an embodiment of this application. Figure 1 As shown, the iris recognition method may include the following steps:

[0033] Step S100: Obtain at least two frames of images to be identified under different supplementary lighting conditions.

[0034] In this embodiment of the application, in order to reduce the impact of light spots on iris recognition, at least two frames of images to be recognized under different supplementary lighting conditions can be acquired. The information complementarity of the images to be recognized under different supplementary lighting conditions can be used to improve the accuracy of iris recognition.

[0035] Step S110: Using a deep learning algorithm, perform feature extraction and feature fusion processing on at least two frames of images to be identified to obtain fused features.

[0036] In this embodiment of the application, in order to achieve information complementarity using images to be identified under different supplementary lighting conditions and avoid the decline in recognition performance due to increased noise after image fusion, feature-level fusion can be used to improve iris recognition performance while achieving information complementarity.

[0037] For example, deep learning algorithms can be used to extract features from at least two frames of images to be identified, and the extracted feature information can be fused to obtain fused features, and high-level semantic fusion can be performed in the feature domain.

[0038] Step S120: Perform iris recognition based on fusion features.

[0039] In this embodiment of the application, when the fusion feature is obtained in the manner described above, iris recognition can be performed based on the fusion feature.

[0040] It can be seen that, in Figure 1 In the method flow shown, at least two frames of images to be identified under different lighting conditions are acquired, and deep learning algorithms are used to extract features and fuse features of the at least two frames of images to be identified. Through high-level semantic fusion of the feature domain, information complementarity of images under different lighting conditions is achieved. While achieving information complementarity, the increase in noise caused by image fusion is avoided. Furthermore, by performing iris recognition based on the fused features, the performance of iris recognition is improved.

[0041] In some embodiments, step S100, acquiring at least two frames of images to be identified under different supplementary lighting conditions, may include:

[0042] The iris camera continuously captures at least two frames of images when supplementary lighting is provided by supplementary lighting devices at different locations; wherein the switching frequency of the supplementary lighting devices is synchronized with the frame rate of the iris camera.

[0043] In this embodiment of the application, in order to reduce the differences between the images to be identified used for feature fusion and optimize the effect of feature fusion, the switching frequency of the supplementary lighting device can be set to be different from the frame rate of the iris camera. That is, the supplementary lighting device is switched once for each frame of image captured by the iris camera, and the supplementary lighting state of adjacent frames captured by the iris camera is inconsistent (supplementary lighting is performed by supplementary lighting devices at different positions).

[0044] Accordingly, in this case (where the switching frequency of the supplementary lighting device is synchronized with the frame rate of the iris camera), at least two frames of images continuously captured by the iris camera can be obtained when supplementary lighting devices are used at different locations. Based on these at least two frames of images continuously captured by the iris camera, feature fusion is performed in the manner described above, which reduces the differences between the images participating in feature fusion and optimizes the effect of feature fusion.

[0045] For example, if N different lighting devices can be used for supplementary lighting, the iris camera can continuously capture N frames of images; N≥2.

[0046] For example, when two different lighting devices are used for supplementary lighting, the iris camera can continuously capture two consecutive frames of images (which can be referred to as odd frame images and even frame images, respectively).

[0047] For example, an iris camera can continuously capture three frames of images when supplementary lighting is provided by three different lighting devices.

[0048] It should be noted that in this embodiment, the more locations where supplementary lighting devices are deployed, the easier it is to obtain complete information when the images from each location are fused. However, the more locations where supplementary lighting devices are deployed, the higher the deployment cost, processing time, and amount of data processed will be. Therefore, the location of the supplementary lighting devices should be selected based on actual needs and desired effects.

[0049] In one example, the supplementary lighting devices at different locations include a first supplementary lighting device at a first supplementary lighting location and a second supplementary lighting device at a second supplementary lighting location;

[0050] The location where the iris information is lost in the first image to be identified is different from the location where the iris information is lost in the second image to be identified;

[0051] The first image to be identified and the second image to be identified are odd-frame and even-frame images captured by the iris camera, respectively; the first image to be identified is the image captured by the iris camera when the first supplementary lighting device provides supplementary lighting, and the second image to be identified is the image captured by the iris camera when the second supplementary lighting device provides supplementary lighting.

[0052] For example, consider deploying supplementary lighting devices at two different locations, i.e., performing information completion by feature fusion on two frames of images.

[0053] For example, supplementary lighting devices (which can be referred to as the first supplementary lighting device and the second supplementary lighting device, respectively) can be deployed at two different locations (which can be referred to as the first location and the second location, respectively). By setting the first location and the second location, the location where the iris information is lost in the image captured by the iris camera (referred to as the first image to be identified in this document) when the first supplementary lighting device is providing supplementary lighting is different from the location where the iris information is lost in the image captured by the iris camera (referred to as the second image to be identified in this document) when the second supplementary lighting device is providing supplementary lighting is different.

[0054] Since the switching frequency of the supplementary lighting device is synchronized with the frame rate of the iris camera, when the first supplementary lighting device is used, the image captured by the iris camera (in the case of non-first frame) is the next frame image captured by the iris camera when the second supplementary lighting device is used (i.e., when the iris camera captures the previous frame image, the second supplementary lighting device provides supplementary lighting); when the iris camera captures the next frame image, it switches back to the second supplementary lighting device for supplementary lighting. That is, the first image to be identified and the second image to be identified are respectively the odd-numbered frame image (referred to as odd frame image in this article) and the even-numbered frame image (referred to as even frame image in this article) in the continuously captured images of the iris camera.

[0055] For example, the first image to be identified is an odd-frame image captured by an iris camera, and the second image to be identified is an even-frame image captured by an iris camera (i.e., the first supplementary lighting device performs supplementary lighting first); or, the first image to be identified is an even-frame image captured by an iris camera, and the second image to be identified is an odd-frame image captured by an iris camera (i.e., the second supplementary lighting device performs supplementary lighting first).

[0056] For example, when the iris camera captures the first frame of the image, the first supplementary lighting device provides supplementary lighting. When the iris camera captures the second frame of the image, the second supplementary lighting device provides supplementary lighting. When the iris camera captures the third frame of the image, the first supplementary lighting device provides supplementary lighting. ... When the iris camera captures the (2n+1)th frame of the image, the first supplementary lighting device provides supplementary lighting. When the iris camera captures the 2nth frame of the image, the second supplementary lighting device provides supplementary lighting.

[0057] It should be noted that, in the embodiments of this application, the difference between the location of the lost iris information in the first image to be identified and the location of the lost iris information in the second image to be identified may include completely different locations of the lost iris information, or the lost iris information areas may have a tolerable overlap, for example, the proportion of the overlapping area of ​​the lost iris information areas to the total area of ​​the lost iris information areas is less than a preset proportion threshold.

[0058] In some embodiments, such as Figure 2 As shown, in step S110, a deep learning algorithm is used to extract and fuse features from at least two frames of images to be identified, resulting in fused features. This can be achieved through the following steps:

[0059] Step S111: Use different feature extraction deep neural networks to extract features from at least two frames of images to be identified;

[0060] Step S112: Using a feature fusion deep neural network, feature fusion processing is performed on the feature information of at least two frames of images to be identified to obtain fused features.

[0061] For example, in order to achieve feature fusion of at least two frames of images to be identified under different lighting conditions, a corresponding deep neural network for feature extraction (referred to as feature extraction deep neural network in this paper) can be trained for the images to be identified under different lighting conditions.

[0062] For example, images can be acquired and labeled under different lighting conditions to obtain training samples under different lighting conditions, which can then be used to train the corresponding feature extraction deep neural network.

[0063] In addition, a deep neural network (referred to as a feature fusion deep neural network in this paper) can be trained to perform feature fusion processing on the feature information extracted by the feature extraction deep neural network.

[0064] Accordingly, for the at least two frames of images to be identified obtained in step S100, different feature extraction deep neural networks can be used to extract features respectively, and feature fusion deep neural networks can be used to perform feature fusion processing on the feature information of the at least two frames of images to be identified to obtain fused features.

[0065] For example, if feature fusion processing is performed on three frames of images to be identified (images to be identified under three different lighting conditions), feature extraction and feature fusion processing can be performed on them. Three different feature fusion deep neural networks can be trained in advance to extract features from the images to be identified under different lighting conditions.

[0066] For example, feature fusion deep neural network 1 is used to extract features from the image to be identified under the first supplementary lighting state, feature fusion deep neural network 2 is used to extract features from the image to be identified under the second supplementary lighting state, and feature fusion deep neural network 3 is used to extract features from the image to be identified under the third supplementary lighting state. The feature information of the image to be identified extracted by feature fusion deep neural network 1, feature fusion deep neural network 2, and feature fusion deep neural network 3 is subjected to feature fusion processing to obtain fused features.

[0067] In some embodiments, step S120, performing iris recognition based on fusion features, may include:

[0068] Iris recognition is performed based on iris comparison thresholds and fusion features;

[0069] The iris matching threshold is adaptively adjusted based on the specified attributes of the image to be identified.

[0070] For example, considering that the degree of occlusion of the iris region in the image and the image quality of the iris region will affect the accuracy of iris recognition, if the same iris comparison threshold is used for iris recognition in cases where the degree of occlusion of the iris region and the image quality of the iris region are different, the iris recognition effect may not be guaranteed.

[0071] For example, if the iris matching threshold is set too high, in scenarios where the iris region is heavily obscured or the image quality of the iris region is poor, the accuracy of iris recognition will be poor, potentially leading to a failure to identify the target based on the iris. Conversely, if the iris matching threshold is set too low, it may increase the probability of false identification.

[0072] Accordingly, in order to improve iris recognition performance, the iris matching threshold can be adaptively adjusted according to the specified attributes of the image, and iris recognition can be performed based on the dynamically adjusted iris matching threshold.

[0073] In one example, the iris matching threshold is determined based on at least one of the following properties of the image to be identified:

[0074] The attributes of wearing glasses, the degree of occlusion, and the overall quality of the iris.

[0075] For example, considering that whether the target to be identified is wearing glasses, whether there is occlusion in the iris region of the image to be identified, and the overall quality of the iris will all affect the iris recognition effect, the iris comparison threshold can be adaptively adjusted based on one or more of the following attributes (which can be called specified attributes): the glasses wearing attribute (i.e., whether the target is wearing glasses), the occlusion attribute, and the overall quality of the iris in the image to be identified. Iris recognition can then be performed based on the determined iris comparison threshold.

[0076] To enable those skilled in the art to better understand the technical solutions provided in the embodiments of this application, the technical solutions provided in the embodiments of this application are described below with reference to specific examples.

[0077] Please see Figure 3 This is a deployment diagram of an iris recognition scenario provided in an embodiment of this application, as shown below. Figure 3 As shown, a supplementary lighting device (such as the first supplementary lighting device and the second supplementary lighting device mentioned above) is deployed at the first supplementary lighting position and the second supplementary lighting position on the left and right sides of the iris camera, and the supplementary lighting is carried out alternately according to the principle of different light spot positions.

[0078] For example, the frame rate of the iris camera and the switching frequency of the supplementary lighting device can be adjusted to synchronize. That is, the supplementary lighting device is switched once for each frame of image captured by the iris camera. In this way, the iris images continuously captured by the iris camera can be obtained for feature fusion and iris recognition.

[0079] For example, features can be fused from two consecutive frames of iris images captured by an iris camera (which can be referred to as odd-frame iris images and even-frame iris images, respectively), and iris recognition can be performed based on the fused features.

[0080] based on Figure 3 The scenario shown is as follows: Figure 4 As shown, the iris recognition implementation process is as follows (taking iris recognition of people wearing glasses as an example):

[0081] 1. By alternately applying supplementary lighting to the face using supplementary lights in different positions (taking supplementary lighting devices as an example), an iris image of the person wearing glasses can be obtained;

[0082] 2. By synchronizing the image frame rate and the alternation frequency of the supplementary light (i.e., the switching frequency of the supplementary light device), iris images (i.e., odd and even frame iris images) under different supplementary light conditions at consecutive close moments are obtained.

[0083] 3. Use deep neural networks (such as convolutional neural networks (CNN)) to perform image fusion at the feature level and extract fused iris features;

[0084] 4. Perform feature comparison under adaptive thresholds based on human eye and iris attributes;

[0085] 5. Output the comparison results.

[0086] For example, in the above process, the overall image information in the odd and even frame iris images is more comprehensive and more reliable.

[0087] Assuming that the iris information lost due to supplemental lighting in odd-numbered iris images is Ia, and the iris information lost due to supplemental lighting in even-numbered iris images is Ib, then the requirements of the overall supplemental lighting scheme can be expressed by the following formula:

[0088] Ia∩Ib=0

[0089] That is, the location where iris information is lost in odd-numbered iris images is different from the location where iris information is lost in even-numbered iris images.

[0090] For example, to address the issue of light spots in iris images of people wearing glasses, a deep neural network can be used to fuse odd and even frame iris images at the feature level.

[0091] For example, different convolutional neural networks (i.e., the aforementioned feature extraction deep neural networks) can be used to extract features from odd-frame iris images and even-frame iris images respectively. Then, a feature fusion convolutional neural network (i.e., the aforementioned feature fusion deep neural network) can be used to perform feature fusion processing on the extracted feature information to improve the robustness of iris features. The implementation process can be as follows: Figure 5 As shown.

[0092] The above solution can overcome at least two drawbacks of traditional image fusion schemes:

[0093] First, the fusion of regions only needs to be completed at the feature level through a trained neural network, without the need for region segmentation and registration of the image, which is simple to implement and does not require coupling between the various steps.

[0094] Second, combining image fusion and feature extraction enables high-level semantic fusion in the feature domain, avoiding increased noise caused by direct image fusion, which in turn leads to a decrease in recognition performance.

[0095] For example, in order to perform iris recognition more accurately, an adaptive threshold feature comparison scheme can be used to achieve iris recognition.

[0096] For example, the iris matching threshold can be adaptively adjusted according to the degree of occlusion of the iris region and the image quality of the iris region, so that iris images with different specified attributes have different iris matching thresholds, thereby improving the overall user experience.

[0097] For example, an adaptive threshold adjustment strategy can be as follows:

[0098] Acquire various attribute data of the iris image, such as the glasses-wearing attribute value Ig, the occlusion attribute value Iz, and the overall iris quality Iq. Then, by comprehensively considering the aforementioned attribute values ​​through a threshold mapping function, obtain the iris matching threshold for that attribute combination state.

[0099] Thd = f(Ig, Iz, Iq, ...)

[0100] For example, iris recognition can be performed based on the iris comparison threshold to identify the identity information of the target to be identified.

[0101] The method provided in this application has been described above. The apparatus provided in this application is described below:

[0102] Please see Figure 6 This is a schematic diagram of the structure of an iris recognition device provided in an embodiment of this application, as shown below. Figure 6 As shown, the iris recognition device may include:

[0103] The acquisition unit 610 is used to acquire at least two frames of images to be identified under different supplementary lighting conditions;

[0104] The feature processing unit 620 is used to perform feature extraction and feature fusion processing on the at least two frames of images to be identified using a deep learning algorithm to obtain fused features;

[0105] The recognition unit 630 is used to perform iris recognition based on the fusion features.

[0106] In some embodiments, the acquisition unit 610 acquires at least two frames of images to be identified under different supplementary lighting conditions, including:

[0107] When supplementary lighting is applied from different locations, the iris camera continuously captures at least two frames of images; wherein the switching frequency of the supplementary lighting device is synchronized with the frame rate of the iris camera.

[0108] In some embodiments, the supplementary lighting devices at different locations include a first supplementary lighting device at a first supplementary lighting location and a second supplementary lighting device at a second supplementary lighting location;

[0109] The location where the iris information is lost in the first image to be identified is different from the location where the iris information is lost in the second image to be identified;

[0110] The first image to be identified and the second image to be identified are odd-frame and even-frame images captured by the iris camera, respectively; the first image to be identified is the image captured by the iris camera when the first supplementary lighting device provides supplementary lighting, and the second image to be identified is the image captured by the iris camera when the second supplementary lighting device provides supplementary lighting.

[0111] In some embodiments, the feature processing unit 620 uses a deep learning algorithm to perform feature extraction and feature fusion processing on the at least two frames of images to be identified, to obtain fused features, including:

[0112] Different feature extraction deep neural networks are used to extract features from the at least two frames of images to be identified.

[0113] By using a feature fusion deep neural network, feature fusion processing is performed on the feature information of the extracted at least two frames of images to be identified to obtain fused features.

[0114] In some embodiments, the identification unit 630 performs iris recognition based on the fusion features, including:

[0115] Iris recognition is performed based on the iris comparison threshold and the fusion features.

[0116] The iris matching threshold is adaptively adjusted based on the specified attributes of the image to be identified.

[0117] In some embodiments, the iris matching threshold is determined based on at least one of the following properties of the image to be identified:

[0118] The attributes of wearing glasses, the degree of occlusion, and the overall quality of the iris.

[0119] This application provides an electronic device including a processor and a memory, wherein the memory stores machine-executable instructions that can be executed by the processor, and the processor executes the machine-executable instructions to implement the iris recognition method described above.

[0120] Please see Figure 7 This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of this application. The electronic device may include a processor 701 and a memory 702 storing machine-executable instructions. The processor 701 and the memory 702 can communicate via a system bus 703. Furthermore, by reading and executing the machine-executable instructions corresponding to the iris recognition logic in the memory 702, the processor 701 can execute the iris recognition method described above.

[0121] The memory 702 mentioned in this document can be any electronic, magnetic, optical, or other physical storage device that can contain or store information such as executable instructions, data, etc. For example, machine-readable storage media can be: RAM (Random Access Memory), volatile memory, non-volatile memory, flash memory, storage drives (such as hard disk drives), solid-state drives, any type of storage disk (such as optical discs, DVDs, etc.), or similar storage media, or combinations thereof.

[0122] In some embodiments, a machine-readable storage medium, such as Figure 7 The memory 702 in the device stores machine-executable instructions, which, when executed by a processor, implement the iris recognition method described above. For example, the storage medium may be ROM, RAM, CD-ROM, magnetic tape, floppy disk, or optical data storage device.

[0123] Please see Figure 8 This is a schematic diagram of the architecture of an iris recognition system provided in an embodiment of this application, as shown below. Figure 8 As shown, the iris recognition system may include: an image acquisition device 810, a supplementary lighting device 820, and an iris recognition device 830; wherein:

[0124] The image acquisition device 810 is used for image acquisition;

[0125] The supplementary lighting device 820 is used for supplementary lighting;

[0126] The iris recognition device 830 is used to perform iris recognition in accordance with the manner described in the above method embodiments.

[0127] It should be noted that, in the embodiments of this application, the image acquisition device 810 and the iris recognition device 830 can be the same device, such as a video front-end device with intelligent analysis function.

[0128] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0129] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. An iris recognition method, characterized in that, include: Acquire at least two frames of images to be identified under different supplementary lighting conditions; Using deep learning algorithms, feature extraction and feature fusion processing are performed on the at least two frames of images to be identified to obtain fused features; Iris recognition is performed based on the fusion features; The step of acquiring at least two frames of images to be identified under different supplementary lighting conditions includes: When supplementary lighting is applied from different locations, the iris camera continuously captures at least two frames of images; wherein the switching frequency of the supplementary lighting device is synchronized with the frame rate of the iris camera. The supplementary lighting devices at different locations include a first supplementary lighting device at a first supplementary lighting location and a second supplementary lighting device at a second supplementary lighting location; The location where the iris information is lost in the first image to be identified is different from the location where the iris information is lost in the second image to be identified; The first image to be identified and the second image to be identified are odd-frame and even-frame images captured by the iris camera, respectively; the first image to be identified is the image captured by the iris camera when the first supplementary lighting device provides supplementary lighting, and the second image to be identified is the image captured by the iris camera when the second supplementary lighting device provides supplementary lighting.

2. The method according to claim 1, characterized in that, The process of using deep learning algorithms to extract and fuse features from at least two frames of images to be identified, resulting in fused features, includes: Different feature extraction deep neural networks are used to extract features from the at least two frames of images to be identified. By using a feature fusion deep neural network, feature fusion processing is performed on the feature information of the extracted at least two frames of images to be identified to obtain fused features.

3. The method according to claim 1, characterized in that, The iris recognition based on the fusion features includes: Iris recognition is performed based on the iris comparison threshold and the fusion features. The iris matching threshold is adaptively adjusted based on the specified attributes of the image to be identified.

4. The method according to claim 3, characterized in that, The iris matching threshold is determined based on at least one of the following attributes of the image to be identified: The attributes of wearing glasses, the degree of occlusion, and the overall quality of the iris.

5. An iris recognition device, characterized in that, include: The acquisition unit is used to acquire at least two frames of images to be identified under different supplementary lighting conditions; The feature processing unit is used to perform feature extraction and feature fusion processing on the at least two frames of images to be identified using deep learning algorithms to obtain fused features; The identification unit is used to perform iris recognition based on the fusion features; The acquisition unit acquires at least two frames of images to be identified under different supplementary lighting conditions, including: When supplementary lighting is applied from different locations, the iris camera continuously captures at least two frames of images; wherein the switching frequency of the supplementary lighting device is synchronized with the frame rate of the iris camera. The supplementary lighting devices at different locations include a first supplementary lighting device at a first supplementary lighting location and a second supplementary lighting device at a second supplementary lighting location; The location where the iris information is lost in the first image to be identified is different from the location where the iris information is lost in the second image to be identified; The first image to be identified and the second image to be identified are odd-frame and even-frame images captured by the iris camera, respectively; the first image to be identified is the image captured by the iris camera when the first supplementary lighting device provides supplementary lighting, and the second image to be identified is the image captured by the iris camera when the second supplementary lighting device provides supplementary lighting.

6. The apparatus according to claim 5, characterized in that, The feature processing unit uses a deep learning algorithm to perform feature extraction and feature fusion processing on the at least two frames of images to be identified, to obtain fused features, including: Different feature extraction deep neural networks are used to extract features from the at least two frames of images to be identified. By using a feature fusion deep neural network, feature fusion processing is performed on the feature information of the extracted at least two frames of images to be identified to obtain fused features; And / or, The recognition unit performs iris recognition based on the fused features, including: Iris recognition is performed based on the iris comparison threshold and the fusion features. The iris matching threshold is adaptively adjusted based on a specified attribute of the image to be identified. The iris matching threshold is determined based on at least one of the following attributes of the image to be identified: The attributes of wearing glasses, the degree of occlusion, and the overall quality of the iris.

7. An electronic device, characterized in that, The method includes a processor and a memory, the memory storing machine-executable instructions that can be executed by the processor, the processor executing the machine-executable instructions to implement the method as described in any one of claims 1-4.

8. An iris recognition system, characterized in that, include: Image acquisition equipment, supplementary lighting equipment, and iris recognition equipment; among which: The image acquisition device is used for image acquisition; The supplemental lighting device is used for providing supplemental lighting; The iris recognition device is used to perform iris recognition in accordance with the manner described in any one of claims 1-4.

Images