Identity authentication method, apparatus and module

Abstract

The embodiments of the present application relate to the technical field of image processing. Provided are an identity authentication method, apparatus and module, which are applied to a processor in an identity authentication module. The method comprises: identifying a first feature region in a first image acquired by an image acquisition assembly; when a first type of a biological feature to be extracted from the first feature region is inconsistent with a second type corresponding to an image processing parameter that has been loaded into a second memory and the first type is a registered type, loading, to the second memory and from a first memory, an image processing parameter corresponding to the first type, so as to replace the image processing parameter corresponding to the second type in the second memory; on the basis of the image processing parameter that has been loaded into the second memory, performing image processing on the first feature region, and extracting a first biological feature from the processed first feature region; and on the basis of the first biological feature and a reference biological feature of the first type, performing identity authentication. By applying the solution provided in the embodiments of the present application, the hardware cost of identity recognition can be reduced.

Description

An identity authentication method, device and module

[0001] This application claims priority to Chinese Patent Application No. 202510192004.2, filed on February 20, 2025, entitled "An Identity Authentication Method, Device and Module", the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to the field of image processing technology, and in particular to an identity authentication method, apparatus and module. Background Technology

[0003] Biometric authentication technology, with its convenience and efficiency, is widely used in various scenarios such as access control and financial payments. Currently, various biometric features, including facial features and palm vein features, can be used for authentication. Taking access control as an example, the authentication module in a smart lock can collect a visitor's facial image and authenticate the visitor's identity based on the facial features extracted from the image. Alternatively, it can collect a visitor's palm vein image and authenticate the visitor's identity based on the palm vein features extracted from the image.

[0004] In other words, in related technologies, identity authentication modules typically support authentication using different types of biometrics. Therefore, identity authentication modules usually deploy multiple independent image acquisition components. These components extract the corresponding type of biometrics from the images they acquire, and then perform identity authentication based on the extracted biometrics. This results in a large number of hardware components in the identity authentication module, leading to higher hardware costs. Summary of the Invention

[0005] The purpose of this application is to provide an identity authentication method, apparatus, and module to reduce the hardware cost of identity recognition. The specific technical solution is as follows:

[0006] In a first aspect, embodiments of this application provide an identity authentication method applied to a processor in an identity authentication module. The identity authentication module further includes: an image acquisition component, a non-volatile first memory, and a volatile second memory. The first memory stores multiple sets of image processing parameters, each set corresponding to a type of biometric feature. The second memory is loaded with one set of image processing parameters from the multiple sets of image processing parameters. The method includes:

[0007] Identify the first feature region in the first image acquired by the image acquisition component;

[0008] If the first type of the biometric feature to be extracted in the first feature region is inconsistent with the second type corresponding to the image processing parameters loaded in the second memory, and the first type is a registered type, then the image processing parameters corresponding to the first type are loaded from the first memory to the second memory to replace the image processing parameters corresponding to the second type in the second memory. The registered type is: the type of the baseline biometric feature that has been recorded.

[0009] Based on the image processing parameters loaded in the second memory, the first feature region is processed to extract the first biometric feature of the processed first feature region.

[0010] Identity authentication is performed based on the first biometric feature and the baseline biometric feature of the first type.

[0011] Secondly, embodiments of this application provide an identity authentication device, applied to a processor in an identity authentication module. The identity authentication module further includes: an image acquisition component, a non-volatile first memory, and a volatile second memory. The first memory stores multiple sets of image processing parameters, each set corresponding to a type of biometric feature. The second memory loads one set of image processing parameters from the multiple sets. The device includes:

[0012] The first recognition module is used to recognize a first feature region in the first image acquired by the image acquisition component;

[0013] The first parameter loading module is configured to load the image processing parameters corresponding to the first type from the first memory to the second memory if the first type of the biometric feature to be extracted in the first feature region is inconsistent with the second type corresponding to the image processing parameters already loaded in the second memory, and the first type is a registered type, so as to replace the image processing parameters corresponding to the second type in the second memory. The registered type is: the type of the baseline biometric feature that has been recorded.

[0014] The first feature extraction module is used to perform image processing on the first feature region based on the image processing parameters loaded in the second memory, and extract the first biometric feature of the processed first feature region.

[0015] The first identity authentication module is used to perform identity authentication based on the first biometric feature and the first type of baseline biometric feature.

[0016] Thirdly, embodiments of this application provide an identity authentication module, including:

[0017] An image acquisition component is used to acquire images and send the acquired images to the processor.

[0018] A non-volatile first memory is used to store multiple sets of image processing parameters, each set of image processing parameters corresponding to a type of biometric feature;

[0019] A volatile second memory is used to load one set of image processing parameters from the plurality of image processing parameters;

[0020] The processor is configured to execute the method described in the first aspect.

[0021] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the method described in the first aspect.

[0022] Fifthly, embodiments of this application also provide a computer program product containing instructions that, when run on a computer, cause the computer to perform the method described in the first aspect.

[0023] As can be seen from the above, in the solution provided by this application embodiment, the relationship between the type of biometric feature to be extracted and the currently loaded image processing parameters is considered before extracting biometric features from the first image. Specifically, the switching of image processing parameters is determined based on whether the first type of the biometric feature to be extracted matches the second type corresponding to the currently loaded image processing parameters. This ensures that the image processing parameters match the biometric feature to be extracted, thereby enabling the extraction of accurate and high-quality biometric features based on the matched image processing parameters. Consequently, the identity authentication module can smoothly perform subsequent identity authentication based on the extracted accurate and high-quality biometric features. In other words, through flexible switching of image processing parameters, the image processing parameters and the biometric feature to be extracted can match under various circumstances. This allows for identity authentication based on various biometric features based on images acquired by a single image acquisition component, eliminating the need for multiple independent image acquisition components and other hardware assistance such as distance sensors. This reduces the number of hardware components in the identity authentication module and lowers the hardware cost of identity recognition.

[0024] Of course, implementing any product or method of this application does not necessarily require achieving all of the advantages described above at the same time. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other embodiments can be obtained based on these drawings.

[0026] Figure 1 is a flowchart illustrating an identity authentication method provided in an embodiment of this application;

[0027] Figure 2 is a flowchart illustrating a biometric data entry method provided in an embodiment of this application;

[0028] Figure 3 is a schematic diagram of a face template registration process provided in an embodiment of this application;

[0029] Figure 4 is a schematic diagram of a hand template registration process provided in an embodiment of this application;

[0030] Figure 5 is a schematic diagram of an identity authentication process provided in an embodiment of this application;

[0031] Figure 6 is a schematic diagram of the structure of an identity authentication device provided in an embodiment of this application;

[0032] Figure 7 is a schematic diagram of the structure of an identity authentication module provided in an embodiment of this application. Detailed Implementation

[0033] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art based on this application are within the scope of protection of this application.

[0034] First, the implementing entity of the solution provided in the embodiments of this application will be described.

[0035] The implementation entity of the solution provided in this application embodiment is the processor in the identity authentication module. The identity authentication module includes an image acquisition component, a non-volatile first memory, a volatile second memory, and a processor.

[0036] The image acquisition component is used to acquire images and send the acquired images to the processor; the non-volatile first memory is used to store multiple sets of image processing parameters, each set of image processing parameters corresponding to a type of biometric feature; the volatile second memory is used to load one set of image processing parameters from the multiple sets of image processing parameters; the processor is used to execute the identity authentication scheme provided in the embodiments of this application based on the received images.

[0037] The aforementioned non-volatile first memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an electrically alterable read-only memory (EAROM), or the like. In one case, the first memory can be flash memory, which is a type of ROM.

[0038] The aforementioned volatile second memory refers to random access memory (RAM) type memory, specifically dynamic random access memory (DRAM), static random access memory (SRAM), and other types of memory.

[0039] The first memory stores multiple sets of image processing parameters that are pre-set, each parameter being used to extract a specific type of biometric feature. In other words, for each set of image processing parameters, processing the image based on that set of parameters allows for the extraction of high-quality, accurate image features.

[0040] The specific meanings of the above-mentioned image processing parameters, biometrics, and other concepts are explained in the specific embodiments below, and will not be detailed here.

[0041] The application scenarios of the solutions provided in the embodiments of this application will be further described.

[0042] The application scenarios of the solution provided in this application can include various biometric authentication scenarios such as access control and mobile payment. For example, in a mobile payment scenario, merchants can authenticate customers' identities based on their biometric features and process payments upon successful authentication.

[0043] The identity authentication scheme provided in the embodiments of this application will now be described in detail with reference to the accompanying drawings.

[0044] Referring to Figure 1, it is a flowchart of an identity authentication method provided in an embodiment of this application. The above method is applied to the processor in the identity authentication module and includes the following steps S101-S104.

[0045] Step S101: Identify the first feature region in the first image acquired by the image acquisition component.

[0046] The aforementioned image acquisition component can be a component for acquiring visible light images, a component for acquiring infrared images, or a component capable of acquiring both visible light and infrared images. Correspondingly, the first image can be either a visible light image or an infrared image.

[0047] Among them, the infrared image acquisition component captures more obvious features such as palm veins in the infrared images, which helps to improve the accuracy of the extracted features for identity authentication; in addition, infrared light imaging is less affected by changes in external light, and can stably acquire images that can be used for identity authentication regardless of whether it is a strong light daytime scene or a weak light nighttime scene.

[0048] In the solution of this application embodiment, identity authentication can be performed by extracting biometric features from the first image.

[0049] The biometric features used for identity authentication may include various types, and this application embodiment does not limit them.

[0050] In one instance, the types of the aforementioned biometrics include at least two of the following types:

[0051] Palms, face, back of hands, eyes, earlobes, fingers.

[0052] Among them, biometric features of the palm type can include palm print features, palm vein features, etc.; biometric features of the back of the hand type can include back of the hand vein features, etc.; biometric features of the eye type can include iris features, etc.; and biometric features of the finger type can include fingerprint features, etc.

[0053] In this way, identity authentication can be performed based on a wide variety of biometric features, further improving the convenience of identity authentication.

[0054] The first feature region can be a region in the first image used to extract the various biometric features mentioned above, such as the palm region, face region, back of hand region, eye region, and finger region in the first image.

[0055] The specific method for identifying the first feature region from the first image is described below.

[0056] In one implementation, a pattern matching algorithm can be used to identify a first feature region in a first image.

[0057] Specifically, the first image can be divided into regions. For each set biometric type, image matching algorithms such as normalized cross-correlation algorithm are used to calculate the similarity between the pre-collected template image of that biometric type and each region in the first image. The region with the highest similarity is determined as the region used to extract the biometric feature of that type.

[0058] For example, for the facial biometric feature type, multiple template images can be collected in advance; then, the similarity between the multiple template images and each region in the first image is calculated to obtain the average similarity between each region and the template image; finally, the region with the highest average similarity in the first image is determined as the region used to extract facial biometric features.

[0059] In another implementation, the first image can be input into a pre-trained feature region recognition model to obtain a first feature region in the first image output by the feature region recognition model. The aforementioned feature region recognition model can be a Convolutional Neural Network (CNN) model, a Transformer network model, etc., and this embodiment of the invention is not limited thereto.

[0060] Step S102: If the first type of the biometric feature to be extracted in the first feature region is inconsistent with the second type corresponding to the image processing parameters loaded in the second memory, and the first type is a registered type, then the image processing parameters corresponding to the first type are loaded from the first memory to the second memory to replace the image processing parameters corresponding to the second type in the second memory.

[0061] First, we will introduce the first type, the second type, and the registered type.

[0062] 1. Type 1

[0063] The first type is the type of biometric feature to be extracted in the first feature region. For example, if the first feature region is the face region, then the first type is face; if the first feature region is the palm region, then the first type is palm.

[0064] 2. Second type

[0065] As described above, the first memory stores multiple sets of image processing parameters, each set corresponding to a type of biometric feature. For example, face type corresponds to face image processing parameters, hand type corresponds to hand image processing parameters, and so on.

[0066] The second memory loads one set of image processing parameters from the multiple sets of image processing parameters mentioned above. The type corresponding to the image processing parameters currently stored in the second memory is the second type mentioned above. That is, at any given time, the second memory will load one set of image processing parameters from the first memory. The type of biometric feature corresponding to the image processing parameters currently stored in the second memory is called the second type. For example, if the type corresponding to the image processing parameters currently loaded in the second memory is face, then the second type is face type; if the type corresponding to the image processing parameters currently loaded in the second memory is palm, then the second type is palm type.

[0067] 3. Registered type

[0068] The registered type is: the type of baseline biometric feature that has been entered.

[0069] The aforementioned baseline biometrics are used for feature comparison and identity authentication based on the comparison results. The identity authentication module can pre-enter these baseline biometrics, and the type of the baseline biometrics is the registered type.

[0070] Therefore, for registered types, identity authentication can be performed based on the baseline biometrics of the registered type. That is, identity authentication is performed by comparing the extracted biometrics with the baseline biometrics.

[0071] The following section will explain the meanings of cases where the first and second types are consistent and inconsistent.

[0072] If the first type and the second type are inconsistent, that is, the type of biometric feature to be extracted in the first feature region does not match the type corresponding to the currently loaded image processing parameters, it may be difficult to extract high-quality biometric features based on the currently loaded image processing parameters. Therefore, it can be preliminarily determined that image parameter switching is necessary.

[0073] Furthermore, if the first type is a registered type, it means that the identity recognition module has pre-entered the baseline biometric features corresponding to the first type, meaning that identity authentication can be performed based on the baseline biometric features of the registered type. Therefore, image parameters can be switched to make the biometric feature type consistent with the image processing parameters, thereby extracting high-quality biometric features for subsequent identity authentication operations.

[0074] When switching image parameters, image processing parameters of the first type are loaded from the first memory to the second memory to replace the image processing parameters of the second type in the second memory. That is, the image processing parameters loaded in the second memory are replaced with the image processing parameters of the first type, and a set of image processing parameters is always loaded in the second memory.

[0075] When the first and second types match, meaning the type of the biometric feature to be extracted in the first feature region matches the type corresponding to the currently loaded image processing parameters, high-quality biometric features can be extracted based on the currently loaded image processing parameters without the need for image parameter switching.

[0076] Step S103: Perform image processing on the first feature region based on the image processing parameters loaded in the second memory, and extract the first biometric feature of the processed first feature region.

[0077] In this step, the image processing parameters loaded in the second memory are the currently loaded image processing parameters. Based on the description of step S102 above, if the first type and the second type are the same, there is no need to replace the image processing parameters in the second memory; if the first type and the second type are different, the image processing parameters in the second memory need to be replaced. Therefore, the image processing parameters loaded in the second memory in this step can be either replaced or unreplaced image processing parameters.

[0078] The image processing parameters will be described in detail below.

[0079] Image processing parameters can include automatic exposure (AE), noise reduction, sharpening, gamma correction, and other parameters. Furthermore, the image processing parameters may differ for different biometric types.

[0080] Specifically, staff can determine the image processing parameters corresponding to the biometric type based on their experience and the characteristics of the biometric type. This will enable them to extract more accurate and higher-quality biometric features from the first feature region after processing it according to the image processing parameters.

[0081] The following explanation uses face type and hand type as examples.

[0082] For example, in face-based identity recognition, the face is generally far from the acquisition component; that is, the face in the image is usually at a distance. Furthermore, facial biometrics mainly rely on the overall features of the facial region, with less emphasis on local contrast. Therefore, the average target brightness in the automatic exposure parameters can be set to a higher value, such as 110 luminance units; noise reduction and sharpening can be turned off to preserve facial details; and brightness parameters can be adjusted to brighten dark areas and darken bright areas to make the facial region more clearly visible.

[0083] For example, when identifying individuals based on palm veins, the palm is typically close to the acquisition component, meaning the palm in the image is usually at close range. Furthermore, palm vein features rely on vein information within the palm, requiring high contrast but less reliance on brightness. Therefore, the average target brightness in the automatic exposure parameters can be set to a lower value, such as 70 luminance units; noise reduction and sharpening can be enabled to highlight vein information; and brightness parameters can be determined to darken dark areas and slightly darken bright areas to enhance the contrast between vein pixels and the palm background pixels.

[0084] It should be noted that the above content is only an example for the purpose of understanding image processing parameters. Image processing parameters corresponding to biometric types can also be set based on other considerations, and this application embodiment does not limit this.

[0085] After processing the first feature region according to the image processing parameters loaded in the second memory, the first biometric feature of the processed first feature region can be extracted according to the image feature extraction strategy corresponding to the first type that has been preset.

[0086] The image feature extraction strategy includes: image feature extraction algorithms and image processing models designed for various biometric types, which will not be elaborated here. Thus, the first biometric feature of the processed first feature region can be extracted according to the image feature extraction algorithm or image processing model corresponding to the first type.

[0087] Step S104: Perform identity authentication based on the first biometric feature and the baseline biometric feature of the first type.

[0088] The similarity between a first biometric feature and a baseline biometric feature of a first type can be calculated, and then identity authentication can be performed based on the obtained similarity.

[0089] Specifically, the system can determine whether there is a target similarity greater than a set similarity threshold in the obtained similarity scores. If so, it indicates that there is a feature with a high similarity to the first biometric feature among the already entered baseline biometric features. This means that the user who entered the baseline biometric features has a high biometric similarity to the currently authenticated user, thus confirming successful identity authentication. Conversely, if there is no target similarity greater than the set similarity threshold in the obtained similarity scores, the identity authentication fails.

[0090] As can be seen from the above, when performing identity authentication using the scheme provided in the embodiments of this application, the first feature region in the first image acquired by the image acquisition component is identified. If the first type of the biometric feature to be extracted in the first feature region is inconsistent with the second type corresponding to the image processing parameters loaded in the second memory, and the first type is a registered type, then the image processing parameters corresponding to the first type are loaded from the first memory to the second memory to replace the image processing parameters corresponding to the second type in the second memory. Then, the first feature region is image processed based on the image processing parameters loaded in the second memory to extract the first biometric feature of the processed first feature region. Consequently, identity authentication can be performed based on the first biometric feature and the reference biometric feature of the first type.

[0091] As can be seen, in the solution provided in this application embodiment, the relationship between the type of biometric feature to be extracted and the currently loaded image processing parameters is considered before extracting biometric features from the first image. Specifically, the switching of image processing parameters is determined based on whether the first type of the biometric feature to be extracted matches the second type corresponding to the currently loaded image processing parameters. This ensures that the image processing parameters match the biometric feature to be extracted, thereby enabling the extraction of accurate and high-quality biometric features based on the matched image processing parameters. Consequently, the identity authentication module can smoothly perform subsequent identity authentication based on the extracted accurate and high-quality biometric features. In other words, through flexible switching of image processing parameters, the image processing parameters and the biometric feature to be extracted can match under various circumstances. This allows for identity authentication based on various biometric features based on images acquired by a single image acquisition component, eliminating the need for multiple independent image acquisition components and other hardware assistance such as distance sensors. This reduces the number of hardware components in the identity authentication module and lowers the hardware cost of identity recognition.

[0092] The above step S102 describes the case where the first type and the second type are inconsistent and the first type is a registered type. Other cases will be described below.

[0093] For cases where the first type and the second type are inconsistent, and the first type is not a registered type:

[0094] Since the first type is not a registered type, meaning there is no baseline biometric feature for the first type, it is impossible to perform identity authentication based on the baseline biometric feature. Therefore, the image parameter switching operation can be skipped, and the authentication process can be terminated directly.

[0095] For cases where the first type and the second type are the same, and the first type is a registered type:

[0096] In this case, it means that the type of biometric feature to be extracted in the first feature region matches the type of the currently loaded image processing parameters, and it can be based on the baseline biometric feature of the first type. Therefore, no parameter switching operation is performed, and subsequent image processing operations are performed directly.

[0097] For cases where the first type and the second type are the same, and the first type is not a registered type:

[0098] Similarly, since the first type is not a registered type, identity authentication cannot be performed based on the baseline biometrics. Therefore, the authentication process can be terminated directly.

[0099] In one embodiment of this application, after successfully loading the image processing parameters corresponding to the first type from the first memory to the second memory, the following steps may be included after step S104:

[0100] The image processing parameters corresponding to the highest priority registered type are loaded from the first memory to the second memory to replace the image processing parameters corresponding to the first type in the second memory.

[0101] The recognition priority for various biometric types can be preset. This recognition priority can be set based on factors such as the difficulty of biometric extraction and the accuracy of biometric authentication. This application does not limit this setting; examples will be provided below.

[0102] Taking facial features and palm vein features as examples, assuming that facial features are easier to extract than palm vein features, and that facial features are more accurate for identity authentication than palm vein features, then facial features can be given a higher priority and palm vein features can be given a lower priority.

[0103] By replacing the image processing parameters with those corresponding to the first type of biometric feature to be extracted, the currently loaded image processing parameters can be promptly replaced with those corresponding to the registered type with the highest recognition priority. This prioritizes loading the image processing parameters with the highest recognition priority. This facilitates the extraction of high-quality biometric features based on the image processing parameters corresponding to the highest recognition priority type, reduces the probability of image processing parameter switching during subsequent identity authentication, and improves the accuracy and efficiency of identity authentication.

[0104] In one embodiment of this application, when identifying the feature region in step S101 above, it can also be implemented based on the following steps:

[0105] Step A: Identify the initial feature regions in the first image acquired by the image acquisition component.

[0106] The initial feature region refers to the region in the first image used to extract the aforementioned biometric features. In this embodiment, before determining the first feature region, the region in the first image used to extract the aforementioned biometric features is referred to as the initial feature region. Specifically, similar to determining the first feature region, the initial feature region can be identified based on a recognition algorithm or a pre-trained model, which will not be elaborated further here.

[0107] Step B: If multiple initial feature regions are identified, then based on the selection factor, determine the initial feature region from the multiple initial feature regions whose type of biometric feature to be extracted belongs to the registered type, and use it as the first feature region.

[0108] The following section describes the methods for determining the first feature region based on different selection factors.

[0109] In one case, the selection factor includes the recognition priority of the type of biometric feature to be extracted in the initial feature region.

[0110] In this case, the initial feature region with the highest priority for identifying the type of biometric feature to be extracted can be determined as the first feature region.

[0111] In another case, the selection factor includes the image quality representation value of the initial feature region.

[0112] The aforementioned image quality characterization values ​​can be obtained based on parameters such as sharpness and brightness of the initial image region. Specifically, these parameters can be converted into image quality characterization values ​​according to a set conversion relationship. This conversion relationship can be a pre-set expression; substituting the sharpness, brightness, and other parameters into the expression yields the image quality characterization values. For example, the expression can be a weighted calculation expression. After substituting the parameters into the expression, a weighted calculation can be performed on each parameter according to a first weight coefficient set for each parameter, and the calculation result is used as the image quality characterization value.

[0113] In this case, the initial feature region with the largest image quality characterization value can be determined as the first feature region.

[0114] In another case, the selection factors include the aforementioned recognition priority and image quality characterization values.

[0115] In this case, the recognition priority and image quality characterization value of each initial feature region can be weighted according to the second weighting coefficient set for recognition priority and image quality characterization value to obtain a weighted score. The initial feature region with the highest weighted score is determined as the first feature region. The specific calculation method will not be elaborated here.

[0116] In this way, the initial feature region with the highest recognition priority or the best image quality can be determined based on the selection factor, which is then used as the first feature region. This facilitates the extraction of high-quality biometric features from the determined first feature region, thereby improving the accuracy of subsequent biometric-based identity authentication.

[0117] In one embodiment of this application, if identity authentication based on the first biometric feature and the first type of baseline biometric feature fails when multiple initial feature regions are identified, the following step C can also be performed:

[0118] Step C: Based on the selection factor, determine the type of the biometric feature to be extracted from the initial feature regions that have never participated in identity authentication. The initial feature regions that belong to the registered types are used as the new first feature regions, and identity authentication is performed based on the new first feature regions.

[0119] In this step, the method for determining the new first feature region based on the selection factor is similar to that in step B above, except that this step selects the first feature region from an initial feature region that has never participated in identity authentication. Furthermore, the method for performing identity authentication based on the new first feature region can be found in the previously described method for determining identity authentication, and will not be repeated here.

[0120] If multiple initial feature regions are identified, and identity authentication fails based on the previously determined first feature region, a new first feature region can be selected from the initial feature regions that have never been involved, and identity authentication can be performed again based on the new first feature region, thereby improving the accuracy and success rate of identity authentication.

[0121] In one embodiment of this application, after successful identity authentication, the target baseline biometrics corresponding to the highest target similarity can be determined, the object identifier corresponding to the target baseline biometrics can be determined, and the object indicated by the object identifier can be identified as the object in the first image.

[0122] The following section, referring to Figure 2, will introduce the method for inputting baseline biometrics.

[0123] Referring to Figure 2, it is a flowchart of a biometric data entry method provided in an embodiment of this application. The method includes the following steps S201-S204.

[0124] Step S201: If the second type and the third type are inconsistent, the image processing parameters corresponding to the third type are loaded from the first memory to the second memory to replace the image processing parameters corresponding to the second type in the second memory.

[0125] The third type mentioned above is the registration type selected by the user, which is the type of biometric feature the user currently wants to register, such as the face, palm, and eyes mentioned earlier.

[0126] The second type is inconsistent with the third type, that is, the currently loaded image processing parameters are inconsistent with the registration type selected by the user. Therefore, an image parameter switching operation can be performed to switch the currently loaded image processing parameters to the image processing parameters corresponding to the registration type selected by the user.

[0127] Step S202: Identify the second feature region of the registered type in the second image acquired by the image acquisition component.

[0128] The method for identifying the second feature region from the second image can be obtained based on the aforementioned method for identifying the first feature region from the first image, with the only difference being the specific image and feature region.

[0129] Step S203: Perform image processing on the second feature region based on the image processing parameters loaded in the second memory.

[0130] The meanings of the image processing parameters are explained in the aforementioned embodiments of the identity authentication method, and will not be repeated here.

[0131] Step S204: Extract the second biometric feature of the processed second feature region as the baseline biometric feature for the registration type, and store the baseline biometric feature.

[0132] The method for extracting the second biometric feature can be obtained based on the aforementioned method for extracting the first biometric region, with the only difference being the specific feature region and biometric feature.

[0133] In this way, after storing the baseline biometrics of the registration type, the feature entry is completed, and identity authentication can be performed based on the baseline biometrics of the registered type.

[0134] As can be seen, in this embodiment, the relationship between the user-selected registration type and the currently loaded image processing parameters is considered before feature input. Specifically, the switching of image processing parameters is determined based on whether the registration type matches the second type corresponding to the currently loaded image processing parameters. This ensures that the image processing parameters match the biometric features to be extracted, enabling accurate and high-quality biometric features to be extracted based on the matched image processing parameters. Consequently, the identity authentication module can store high-quality baseline biometric features and successfully complete the input of basic biometric features. In other words, through flexible switching of image processing parameters, the image processing parameters and the biometric features to be extracted can match under various circumstances. This allows for the input of baseline biometric features for various biometric features based on images acquired by a single image acquisition component, eliminating the need for multiple independent image acquisition components and other hardware assistance such as distance sensors. This reduces the number of hardware components in the identity authentication module and lowers the hardware cost of identity recognition.

[0135] In one embodiment of this application, after successfully loading the image processing parameters corresponding to the third type from the first memory to the second memory, the following steps can be performed after storing the baseline biometric features:

[0136] The image processing parameters corresponding to the highest priority registered type are loaded from the first memory to the second memory to replace the image processing parameters corresponding to the third type in the second memory.

[0137] In this way, after replacing the image processing parameters with those corresponding to the registered type, the currently loaded image processing parameters can be promptly replaced with those corresponding to the registered type with the highest recognition priority. This prioritizes loading the image processing parameters with the highest recognition priority, which is beneficial for extracting and inputting high-quality biometric features based on these parameters. It also reduces the probability of image processing parameter switching during subsequent feature input, thereby improving the quality and efficiency of the input features.

[0138] In one embodiment of this application, after successfully loading the image processing parameters corresponding to the third type from the first memory to the second memory, the following step D1 can be performed after storing the baseline biometric features:

[0139] Step D1: If the second feature region is not identified and a registered type exists, load the image processing parameters corresponding to the registered type with the highest identification priority from the first memory to the second memory to replace the image processing parameters corresponding to the third type in the second memory.

[0140] In this case, it indicates that feature entry failed and a registered type already exists. The image processing parameters can be replaced with the image processing parameters corresponding to the registered type with the highest recognition priority. This facilitates the subsequent extraction and entry of high-quality biometric features based on the image processing parameters corresponding to the registered type with the highest recognition priority, and reduces the probability of image processing parameter switching during subsequent feature entry.

[0141] In another embodiment of this application, if the image processing parameters corresponding to the third type are successfully loaded from the first memory to the second memory, the following step D2 can be performed after storing the baseline biometric features:

[0142] Step D2: If the second feature region is not identified and there is no registered type, load the image processing parameters corresponding to the highest priority type from the first memory to the second memory to replace the image processing parameters corresponding to the third type in the second memory.

[0143] The difference between this step and step D1 is that the image processing parameters corresponding to the highest priority type among the registered types are replaced with the image processing parameters corresponding to the highest priority type among all types.

[0144] In this case, it indicates that feature entry failed and no registered type exists. Therefore, the image processing parameters can be replaced with the image processing parameters corresponding to the type with the highest recognition priority among all types. This facilitates the subsequent extraction and entry of high-quality biometric features based on the image processing parameters corresponding to the type with the highest recognition priority, and reduces the probability of image processing parameter switching during subsequent feature entry.

[0145] As described above, under specific circumstances, the solution provided in this application will trigger an image processing parameter switching operation to switch the image processing parameters. This switching can be performed by the Image Signal Processor (ISP) module in the authentication process, specifically including the ISP module actively switching parameters and the ISP module passively switching parameters upon receiving an instruction from the processor.

[0146] The following section uses biometric types, including face and hand types, as examples to explain in detail various scenarios involving active and passive switching of image processing parameters. Notably, face type recognition has a higher priority than hand type recognition.

[0147] Active switching scenario:

[0148] Active switching scenario 1: The user registration template includes both a face template and a hand template. After the user completes a palm vein recognition, the currently loaded ISP parameters are the hand ISP parameters. After the recognition is completed, regardless of the result, the ISP module will actively trigger a parameter switch, changing the hand ISP parameters to the default face ISP parameters.

[0149] Among them, the registration template refers to the aforementioned input baseline features, the face template is the input facial baseline biometric feature, and the palm template is the input palm vein baseline biometric feature; the palm ISP parameter refers to the image processing parameters corresponding to the palm vein, and the face ISP parameter refers to the image processing parameters corresponding to the face.

[0150] Face type recognition has a higher priority than hand type recognition; therefore, the face ISP parameter can also be called the default ISP parameter.

[0151] Scenario 2: The user registration template only contains a face template. After a user completes registration using a hand template, the currently loaded ISP parameters are the hand ISP parameters. After registration, regardless of the registration result, the ISP module will proactively trigger a parameter switch, changing the hand ISP parameters to the default face ISP parameters.

[0152] Active switching scenario 3: The user registration template only contains a hand template. After the user completes a face template registration, the currently loaded ISP parameters are the face ISP parameters. If registration fails, since the ISP parameters are currently the face ISP parameters, but the registration template only contains a hand template, the ISP module will actively trigger a parameter switch to change the face ISP parameters to the hand ISP parameters.

[0153] Active switching scenario 4: The user has not registered a template. The user has completed a palm template registration, and the currently loaded ISP parameters are the palm ISP parameters. If registration fails, since there are no registered templates at this time, and the current ISP parameters are the palm ISP parameters, not the default face ISP parameters, the ISP module will actively trigger a parameter switch to change the palm ISP parameters to the face ISP parameters.

[0154] Active switching scenario 5: The user registration template includes both a face template and a hand template. After a user completes a palm vein registration, the currently loaded ISP parameters are the hand ISP parameters. After registration, regardless of the registration result, the ISP module will actively trigger a parameter switch, changing the hand ISP parameters to the default face ISP parameters.

[0155] Passive switching scenario:

[0156] Passive switching scenario 1: During the template registration phase, if the template type registered by the user is inconsistent with the current ISP parameter type, the processor will notify the ISP module to switch parameters, and the ISP module will passively switch the current ISP parameters to another ISP parameter.

[0157] For example, if a user registers a face template but the current ISP parameter is the palm ISP parameter, the ISP module will receive an instruction from the processor and passively switch the current ISP parameter to the face ISP parameter; similarly, if a user registers a palm template but the current ISP parameter is the face ISP parameter, the ISP module will receive an instruction from the processor and passively switch the current ISP parameter to the palm ISP parameter.

[0158] Passive switching scenario 2: During the identity authentication phase, the feature type corresponding to the identified feature region is inconsistent with the current ISP parameter type. For example, if the detected feature region is a face and the corresponding feature type is face type, but the current ISP parameter is the palm ISP parameter, the ISP module will receive an instruction from the processor and passively switch the current ISP parameter to the face ISP parameter.

[0159] The following section, using Figures 3 and 4 as examples, describes the complete process of user registration for face and hand templates, taking biometric types including face and palm as examples.

[0160] First, refer to Figure 3 for an introduction to the process of registering a face template during template registration.

[0161] The process of registering a face template for a user can be divided into the following two scenarios:

[0162] 1. For users registering for the first time, registering only their face, or registering both their face and hand:

[0163] Since face recognition has a higher priority than hand recognition, the default ISP parameters are face ISP parameters in these cases. The currently registered biometric category is face, meaning that the current ISP parameters match the registered biometric category.

[0164] In other words, high-quality biometric features can be extracted based on the currently loaded ISP parameters. Therefore, face detection can be performed directly to determine whether a face is detected. If so, facial features are extracted and stored, resulting in successful face template registration. Since the registered templates include the face template, no parameter switching is required, and the face ISP parameters are maintained. If not, it indicates that face template registration has failed. In this case, the registered templates either include the face template or there are no registered modules, so the face ISP parameters are also maintained.

[0165] 2. For users who have only registered with the app:

[0166] If a user has only registered their palm, the default ISP parameters currently loaded in memory are the palm ISP parameters, while the currently registered biometric category is face. This means the current ISP parameters do not match the registered biometric category, making it difficult to extract high-quality biometrics based on the currently loaded ISP parameters. Therefore, the palm ISP parameters can be switched to the face ISP parameters (corresponding to passive switching scenario 1).

[0167] Then, face detection is performed to determine whether a face is detected. If yes, facial features are extracted and stored, thus the face template is successfully registered. At this point, the face has been registered, so the face ISP parameters are maintained. If no, it means that the face template registration has failed. At this point, the only registered biometric type is still the palm. Therefore, the face ISP parameters need to be switched to the palm ISP parameters (corresponding to active switching case 3).

[0168] Referring to Figure 4, the process of registering a hand template is described.

[0169] The process of registering a palm template for a user can be divided into the following three scenarios:

[0170] 1. For users registering for the first time:

[0171] Because face recognition has a higher priority than hand recognition, the default ISP parameters are face ISP parameters when no type has been registered. However, the currently registered biometric category is hand, meaning the current ISP parameters do not match the registered biometric category. This means it is difficult to extract high-quality biometric features based on the currently loaded ISP parameters. Therefore, the face ISP parameters can be switched to hand ISP parameters (corresponding to passive switching scenario 1).

[0172] Next, palm detection is performed to determine whether a palm is detected. If it is detected, palm vein features are extracted and stored, thus the palm template is successfully registered. At this time, only the palm template is registered, so no parameter switching is performed, and the palm ISP parameters are maintained. If it is not detected, it means that the palm template registration has failed. At this time, there is no registered biometric type. Therefore, the palm ISP parameters are switched back to the default face ISP parameters (corresponding to active switching case 4).

[0173] 2. For users who have only registered their face, or have registered both their face and hand:

[0174] Because face recognition has a higher priority than hand recognition, the default ISP parameters are always face ISP parameters once a face has been registered. However, the currently registered biometric category is hand, meaning the current ISP parameters do not match the registered biometric category. This means it's difficult to extract high-quality biometric features based on the currently loaded ISP parameters. Therefore, the face ISP parameters can be switched to hand ISP parameters (corresponding to passive switching scenario 1).

[0175] Next, palm detection is performed to determine whether a palm is detected. If it is, palm vein features are extracted and stored, thus the palm template is successfully registered. However, since a face template has been registered, the palm ISP parameters need to be switched to the default face ISP parameters (corresponding to active switching cases 2 and 5). If not, it means that the palm template registration has failed. Again, since a face template has been registered, the palm ISP parameters need to be switched to the default face ISP parameters (corresponding to active switching cases 2 and 5).

[0176] 3. For users who have only registered with the app:

[0177] In this case, since only the palm has been registered, the currently loaded ISP parameters are the palm ISP parameters, and the currently registered biometric category is the palm. This means the current ISP parameters match the registered biometric category, resulting in high-quality biometric extraction based on the currently loaded ISP parameters.

[0178] Therefore, palm detection can be performed directly to determine whether a palm is detected. If a palm is detected, palm vein features are extracted and stored, thus the palm template is successfully registered. Since only the palm has been registered, no switching is performed, and the palm ISP parameters are maintained. If no palm is detected, it means that the palm template registration has failed. Since only the palm has been registered, no switching is performed, and the facial ISP parameters are maintained.

[0179] Referring to Figure 5, the user authentication process will be described, which can be divided into the following three cases:

[0180] 1. Case where only the face has been registered

[0181] Directly perform face detection to determine if a face is detected. If a face is detected, perform facial feature extraction and determine if the feature comparison passes. If yes, the authentication is successful; if no, the authentication fails. If no face is detected, the authentication fails directly.

[0182] The process involves determining whether the feature comparison passes, which means calculating the similarity between the extracted facial features and the registered facial template features, and then performing identity authentication based on the obtained similarity. In short, whether the feature comparison passes is equivalent to whether the identity authentication passes.

[0183] Specifically, it can be determined whether there is a target similarity greater than the set similarity threshold among the obtained similarities. If so, it indicates that the identity authentication is successful; otherwise, it indicates that the identity authentication has failed.

[0184] 2. Case where only the palm has been registered

[0185] The system directly detects the palm and determines whether it is detected. If the palm is detected, the system extracts the palm vein features and determines whether the feature comparison passes. If it passes, the authentication is successful; if it fails, the authentication fails. If the palm is not detected, the authentication fails directly.

[0186] 3. Cases where face and palm images have been registered.

[0187] First, a face detection is performed to determine if a face has been detected:

[0188] If a face is detected, facial features are extracted, and the feature comparison is checked. If the feature comparison passes, the authentication is successful; otherwise, the authentication fails.

[0189] If no face is detected, hand detection is performed. If a hand is detected, the current default face ISP parameters are switched to the hand ISP parameters (corresponding to passive switching case 2). Then, palm vein features are extracted and compared. If the feature comparison passes, authentication is confirmed as successful; otherwise, authentication is confirmed as failed. Regardless of whether the feature comparison passes, the hand ISP parameters are switched back to the default face ISP parameters (corresponding to active switching case 1). If no hand is detected, authentication is confirmed as failed.

[0190] As can be seen from the above, the solution provided in this application embodiment can achieve identity authentication based on multiple biometric features such as face and palm vein using only one image acquisition component (lens). It eliminates the need for multiple image acquisition components and additional hardware such as distance sensors, significantly reducing hardware resource consumption and saving hardware costs. While minimizing costs, it implements multiple biometric authentication functions on the same identity authentication module and obtains different image effects for different targets. Specifically, a common infrared lens (IR lens) can be used to achieve palm vein recognition at 15-35cm and face recognition at 40-120cm, without requiring other hardware assistance.

[0191] Furthermore, by using face detection and hand detection algorithms to accurately identify objects in front of the lens, the ISP parameters can be switched appropriately, making the adjustment of the ISP parameters more precise. Based on the ISP parameter processing, it is easier to extract high-quality features from the image.

[0192] Furthermore, this invention precisely sets ISP parameter settings and switching processes for various situations based on different user registration templates, making the parameter switching process design of the identity authentication module more reasonable and reducing switching losses during the registration and identity authentication stages.

[0193] In the solutions provided in this application, the collection, storage, use, processing, transmission, provision, and disclosure of user personal information are all carried out with the user's knowledge and authorization, and comply with relevant laws and regulations, and do not violate public order and good morals.

[0194] It should be noted that the two-dimensional face images in this embodiment are from a publicly available dataset.

[0195] Corresponding to the above-described identity authentication method, this application also provides an identity authentication device.

[0196] Referring to Figure 6, which is a schematic diagram of the structure of an identity authentication device provided in an embodiment of this application, the device is a processor applied in an identity authentication module. The identity authentication module further includes: an image acquisition component, a non-volatile first memory, and a volatile second memory. The first memory stores multiple sets of image processing parameters, each set of image processing parameters corresponding to a type of biometric feature. The second memory is loaded with one set of image processing parameters from the multiple sets of image processing parameters. The device includes the following modules:

[0197] The first identification module 601 is used to identify a first feature region in the first image acquired by the image acquisition component;

[0198] The first parameter loading module 602 is used to load the image processing parameters corresponding to the first type from the first memory to the second memory if the first type of the biometric feature to be extracted in the first feature region is inconsistent with the second type corresponding to the image processing parameters already loaded in the second memory, and the first type is a registered type. This is to replace the image processing parameters corresponding to the second type in the second memory. The registered type is the type of the baseline biometric feature that has been recorded.

[0199] The first feature extraction module 603 is used to perform image processing on the first feature region based on the image processing parameters loaded in the second memory, and extract the first biometric feature of the processed first feature region.

[0200] The first identity authentication module 604 is used to perform identity authentication based on the first biometric feature and the first type of baseline biometric feature.

[0201] As can be seen from the above, when performing identity authentication using the scheme provided in the embodiments of this application, the first feature region in the first image acquired by the image acquisition component is identified. If the first type of the biometric feature to be extracted in the first feature region is inconsistent with the second type corresponding to the image processing parameters loaded in the second memory, and the first type is a registered type, then the image processing parameters corresponding to the first type are loaded from the first memory to the second memory to replace the image processing parameters corresponding to the second type in the second memory. Then, the first feature region is image processed based on the image processing parameters loaded in the second memory to extract the first biometric feature of the processed first feature region. Consequently, identity authentication can be performed based on the first biometric feature and the reference biometric feature of the first type.

[0202] As can be seen, in the solution provided in this application embodiment, the relationship between the type of biometric feature to be extracted and the currently loaded image processing parameters is considered before extracting biometric features from the first image. Specifically, the switching of image processing parameters is determined based on whether the first type of the biometric feature to be extracted matches the second type corresponding to the currently loaded image processing parameters. This ensures that the image processing parameters match the biometric feature to be extracted, thereby enabling the extraction of accurate and high-quality biometric features based on the matched image processing parameters. Consequently, the identity authentication module can smoothly perform subsequent identity authentication based on the extracted accurate and high-quality biometric features. In other words, through flexible switching of image processing parameters, the image processing parameters and the biometric feature to be extracted can match under various circumstances. This allows for identity authentication based on various biometric features based on images acquired by a single image acquisition component, eliminating the need for multiple independent image acquisition components and other hardware assistance such as distance sensors. This reduces the number of hardware components in the identity authentication module and lowers the hardware cost of identity recognition.

[0203] In one embodiment of this application, if the image processing parameters corresponding to the first type are successfully loaded from the first memory to the second memory, the apparatus further includes:

[0204] The second parameter loading module is used to load the image processing parameters corresponding to the highest priority registered type from the first memory to the second memory after identity authentication is completed based on the first biometric feature and the benchmark biometric feature of the first type, so as to replace the image processing parameters corresponding to the first type in the second memory.

[0205] By replacing the image processing parameters with those corresponding to the first type of biometric feature to be extracted, the currently loaded image processing parameters can be promptly replaced with those corresponding to the registered type with the highest recognition priority. This prioritizes loading the image processing parameters with the highest recognition priority. This facilitates the extraction of high-quality biometric features based on the image processing parameters corresponding to the highest recognition priority type, reduces the probability of image processing parameter switching during subsequent identity authentication, and improves the accuracy and efficiency of identity authentication.

[0206] In one embodiment of this application, the first identification module 601 is specifically used to identify initial feature regions in the first image acquired by the image acquisition component; if multiple initial feature regions are identified, then based on a selection factor, an initial feature region whose type of biometric feature to be extracted belongs to a registered type is determined from the multiple initial feature regions as the first feature region, wherein the selection factor includes: the identification priority of the type of biometric feature to be extracted in the initial feature region, and / or, the image quality characterization value of the initial feature region.

[0207] In this way, the initial feature region with the highest recognition priority or the best image quality can be determined based on the selection factor, which is then used as the first feature region. This facilitates the extraction of high-quality biometric features from the determined first feature region, thereby improving the accuracy of subsequent biometric-based identity authentication.

[0208] In one embodiment of this application, if identity authentication based on the first biometric feature and the first type of baseline biometric feature fails, the device further includes:

[0209] The second identity authentication module is used to determine, based on the selection factor, an initial feature region in the initial feature region that the type of the biometric feature to be extracted belongs to the registered type, and use it as a new first feature region for identity authentication.

[0210] If multiple initial feature regions are identified, and identity authentication fails based on the previously determined first feature region, a new first feature region can be selected from the initial feature regions that have never been involved, and identity authentication can be performed again based on the new first feature region, thereby improving the accuracy and success rate of identity authentication.

[0211] In one embodiment of this application, the baseline biometrics are recorded in the following manner:

[0212] If the second type and the third type are inconsistent, the image processing parameters corresponding to the third type are loaded from the first memory to the second memory to replace the image processing parameters corresponding to the second type in the second memory. The third type is: the registration type selected by the user; identifying the second feature region of the registration type in the second image acquired by the image acquisition component; performing image processing on the second feature region based on the image processing parameters loaded in the second memory; extracting the second biometric feature of the processed second feature region as the baseline biometric feature of the registration type, and storing the baseline biometric feature.

[0213] As can be seen, in this embodiment, the relationship between the user-selected registration type and the currently loaded image processing parameters is considered before feature input. Specifically, the switching of image processing parameters is determined based on whether the registration type matches the second type corresponding to the currently loaded image processing parameters. This ensures that the image processing parameters match the biometric features to be extracted, enabling accurate and high-quality biometric features to be extracted based on the matched image processing parameters. Consequently, the identity authentication module can store high-quality baseline biometric features and successfully complete the input of basic biometric features. In other words, through flexible switching of image processing parameters, the image processing parameters and the biometric features to be extracted can match under various circumstances. This allows for the input of baseline biometric features for various biometric features based on images acquired by a single image acquisition component, eliminating the need for multiple independent image acquisition components and other hardware assistance such as distance sensors. This reduces the number of hardware components in the identity authentication module and lowers the hardware cost of identity recognition.

[0214] In one embodiment of this application, if the image processing parameters corresponding to the third type are successfully loaded from the first memory to the second memory, the apparatus further includes:

[0215] The third parameter loading module is used to store the baseline biometric features and then load the image processing parameters corresponding to the highest priority registered type from the first memory to the second memory, so as to replace the image processing parameters corresponding to the third type in the second memory.

[0216] After replacing the image processing parameters with those corresponding to the registered type, the currently loaded image processing parameters can be promptly replaced with those corresponding to the registered type with the highest recognition priority, meaning that the image processing parameters with the highest recognition priority are loaded first. This facilitates the subsequent extraction and input of high-quality biometric features based on the image processing parameters with the highest recognition priority, reduces the probability of image processing parameter switching during subsequent feature input, and improves the quality and efficiency of the input features.

[0217] In one embodiment of this application, if the image processing parameters corresponding to the third type are successfully loaded from the first memory to the second memory, the apparatus further includes:

[0218] The fourth parameter loading module is used to load the image processing parameters corresponding to the highest priority registered type from the first memory to the second memory when the second feature region is not identified and a registered type exists, so as to replace the image processing parameters corresponding to the third type in the second memory.

[0219] and / or

[0220] The fifth parameter loading module is used to load the image processing parameters corresponding to the highest priority type from the first memory to the second memory when no second feature region is identified and no registered type exists, so as to replace the image processing parameters corresponding to the third type in the second memory.

[0221] In this case, if feature entry fails and a registered type already exists, the image processing parameters can be replaced with the image processing parameters corresponding to the registered type with the highest recognition priority. This is beneficial for extracting and entering high-quality biometric features based on the image processing parameters corresponding to the highest recognition priority type, and reduces the probability of image processing parameter switching during subsequent feature entry. Alternatively, if feature entry fails and no registered type exists, the image processing parameters can be replaced with the image processing parameters corresponding to the highest recognition priority type among all types. This is beneficial for extracting and entering high-quality biometric features based on the image processing parameters corresponding to the highest recognition priority type, and reduces the probability of image processing parameter switching during subsequent feature entry.

[0222] In one embodiment of this application, the image acquisition component is an infrared image acquisition component.

[0223] The infrared image acquisition component captures infrared images with more obvious features such as palm veins, which helps improve the accuracy of identity authentication based on the extracted features. In addition, infrared light imaging is less affected by changes in external light, and can stably acquire images in both bright daylight and low-light nighttime scenarios, thus improving the stability of the solution.

[0224] In one embodiment of this application, the biometrics includes at least two of the following types:

[0225] Palms, face, back of hands, eyes, earlobes, fingers.

[0226] In this way, identity authentication can be performed based on a wide variety of biometric features, further improving the convenience of identity authentication.

[0227] This application embodiment also provides an identity authentication module, as shown in Figure 7, including:

[0228] Image acquisition component 701 is used to acquire images and send the acquired images to processor 704;

[0229] The first memory 702 is used to store multiple sets of image processing parameters, each set of image processing parameters corresponding to a type of biometric feature;

[0230] The second memory 703 is used to load one set of image processing parameters from the plurality of image processing parameters;

[0231] Processor 704 is used to execute the aforementioned authentication method.

[0232] The first memory 702 is a non-volatile memory, the second memory 703 is a volatile memory, and the above-mentioned identity authentication module may also include a communication bus and / or a communication interface. The processor 704, the communication interface, the first memory 702, and the second memory 703 communicate with each other through the communication bus.

[0233] The communication bus mentioned in the aforementioned authentication module can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not indicate that there is only one bus or one type of bus.

[0234] The communication interface is used for communication between the aforementioned authentication module and other devices.

[0235] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0236] In one embodiment of this application, the image acquisition component is an infrared image acquisition component.

[0237] In one embodiment of this application, the module further includes an infrared lamp, which is communicatively connected to the image acquisition component and is used to emit infrared light in response to a lighting command sent by the image acquisition component.

[0238] In another embodiment provided in this application, a computer-readable storage medium is also provided, which stores a computer program that, when executed by a processor, implements the steps of the above-described authentication method.

[0239] In another embodiment provided in this application, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to perform any of the authentication methods described above.

[0240] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a solid-state drive (SSD), etc.

[0241] 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.

[0242] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the device, module, and storage medium embodiments are basically similar to the method embodiments, so the descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0243] The above description is merely a preferred embodiment of this application and is not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application are included within the scope of protection of this application.

Claims

1. An identity authentication method, characterized by, A processor is used in an identity authentication module, the identity authentication module further comprising: an image acquisition component, a non-volatile first memory, and a volatile second memory, wherein the first memory stores multiple sets of image processing parameters, each set of image processing parameters corresponding to a type of biometric feature, and the second memory is loaded with one set of image processing parameters from the multiple sets of image processing parameters; the method includes: Identify the first feature region in the first image acquired by the image acquisition component; If the first type of the biometric feature to be extracted in the first feature region is inconsistent with the second type corresponding to the image processing parameters loaded in the second memory, and the first type is a registered type, then the image processing parameters corresponding to the first type are loaded from the first memory to the second memory to replace the image processing parameters corresponding to the second type in the second memory. The registered type is: the type of the baseline biometric feature that has been recorded. Based on the image processing parameters loaded in the second memory, the first feature region is processed to extract the first biometric feature of the processed first feature region. Identity authentication is performed based on the first biometric feature and the baseline biometric feature of the first type.

2. The method of claim 1, wherein, If the image processing parameters corresponding to the first type are successfully loaded from the first memory to the second memory, the method further includes: After identity authentication is completed based on the first biometric feature and the baseline biometric feature of the first type, the image processing parameters corresponding to the registered type with the highest recognition priority are loaded from the first memory to the second memory to replace the image processing parameters corresponding to the first type in the second memory.

3. The method of claim 1, wherein, The identification of the first feature region in the first image acquired by the image acquisition component includes: Identify the initial feature region in the first image acquired by the image acquisition component; If multiple initial feature regions are identified, then based on a selection factor, an initial feature region whose type of biometric feature to be extracted belongs to a registered type is determined from the multiple initial feature regions and designated as the first feature region. The selection factor includes: the identification priority of the type of biometric feature to be extracted in the initial feature region, and / or, the image quality characterization value of the initial feature region.

4. The method of claim 3, wherein, If identity authentication based on the first biometric feature and the first type of baseline biometric feature fails, the method further includes: Based on the selection factor, an initial feature region in which the type of biometric feature to be extracted belongs to a registered type is determined from the initial feature region that has never participated in identity authentication. This initial feature region is then used as a new first feature region, and identity authentication is performed based on the new first feature region.

5. The method of claim 1, wherein, The baseline biometrics are entered in the following manner: If the second type and the third type are inconsistent, the image processing parameters corresponding to the third type are loaded from the first memory to the second memory to replace the image processing parameters corresponding to the second type in the second memory. The third type is the registration type selected by the user. Identify the second feature region of the registration type in the second image acquired by the image acquisition component; The second feature region is processed based on the image processing parameters loaded in the second memory. The second biometric feature of the processed second feature region is extracted and used as the baseline biometric feature of the registration type, and the baseline biometric feature is stored.

6. The method of claim 5, wherein, If the image processing parameters corresponding to the third type are successfully loaded from the first memory to the second memory, the method further includes: After storing the baseline biometrics, the image processing parameters corresponding to the highest priority registered type are loaded from the first memory to the second memory to replace the image processing parameters corresponding to the third type in the second memory.

7. The method of claim 5, wherein, If the image processing parameters corresponding to the third type are successfully loaded from the first memory to the second memory, the method further includes: If the second feature region is not identified and a registered type exists, the image processing parameters corresponding to the registered type with the highest identification priority are loaded from the first memory to the second memory to replace the image processing parameters corresponding to the third type in the second memory; and / or, if the second feature region is not identified and no registered type exists, the image processing parameters corresponding to the type with the highest priority are loaded from the first memory to the second memory to replace the image processing parameters corresponding to the third type in the second memory.

8. The method according to any one of claims 1-7, characterized in that, The image acquisition component is an infrared image acquisition component; and / or, the biometric type includes at least two of the following types: palm, face, back of hand, eye, auricle, and finger.

9. An identity authentication apparatus characterized by comprising: A processor is used in an identity authentication module, the identity authentication module further comprising: an image acquisition component, a non-volatile first memory, and a volatile second memory, wherein the first memory stores multiple sets of image processing parameters, each set of image processing parameters corresponding to a type of biometric feature, and the second memory is loaded with one set of image processing parameters from the multiple sets of image processing parameters; the device comprises: The first recognition module is used to recognize a first feature region in the first image acquired by the image acquisition component; The first parameter loading module is configured to load the image processing parameters corresponding to the first type from the first memory to the second memory if the first type of the biometric feature to be extracted in the first feature region is inconsistent with the second type corresponding to the image processing parameters already loaded in the second memory, and the first type is a registered type, so as to replace the image processing parameters corresponding to the second type in the second memory. The registered type is: the type of the baseline biometric feature that has been recorded. The first feature extraction module is used to perform image processing on the first feature region based on the image processing parameters loaded in the second memory, and extract the first biometric feature of the processed first feature region. The first identity authentication module is used to perform identity authentication based on the first biometric feature and the first type of baseline biometric feature.

10. The apparatus according to claim 9, characterized in that, If the image processing parameters corresponding to the first type are successfully loaded from the first memory to the second memory, the device further includes: a second parameter loading module, configured to, after identity authentication is completed based on the first biometric feature and the baseline biometric feature of the first type, load the image processing parameters corresponding to the registered type with the highest recognition priority from the first memory to the second memory, so as to replace the image processing parameters corresponding to the first type in the second memory; and / or, the first recognition module is specifically configured to recognize the initial feature regions in the first image acquired by the image acquisition component; if multiple initial feature regions are recognized, then based on a selection factor, determine from the multiple initial feature regions whether the type of biometric feature to be extracted belongs to the registered type. The initial feature region of the registration type is used as the first feature region, wherein the selection factor includes: the recognition priority of the type of biometric feature to be extracted in the initial feature region, and / or, the image quality characterization value of the initial feature region; and / or, if identity authentication based on the first biometric feature and the reference biometric feature of the first type fails, the device further includes: a second identity authentication module, used to determine, based on the selection factor, the type of biometric feature to be extracted from the initial feature region that has never participated in identity authentication belongs to the initial feature region of the registered type, as a new first feature region, and to perform identity authentication based on the new first feature region; and / or, the reference biometric feature is entered in the following manner: if the second type and the third type... If the types are inconsistent, the image processing parameters corresponding to the third type are loaded from the first memory to the second memory to replace the image processing parameters corresponding to the second type in the second memory. The third type is: the registration type selected by the user; identifying the second feature region of the registration type in the second image acquired by the image acquisition component; performing image processing on the second feature region based on the image processing parameters loaded in the second memory; extracting the second biometric feature of the processed second feature region as the baseline biometric feature of the registration type, and storing the baseline biometric feature; and / or, if the image processing parameters corresponding to the third type are successfully loaded from the first memory to the second memory, the assembly... The device further includes: a third parameter loading module, used to load the image processing parameters corresponding to the highest priority registered type from the first memory to the second memory after storing the baseline biometric features, so as to replace the image processing parameters corresponding to the third type in the second memory; and / or, if the image processing parameters corresponding to the third type are successfully loaded from the first memory to the second memory, the device further includes: a fourth parameter loading module, used to load the image processing parameters corresponding to the highest priority registered type from the first memory to the second memory when no second feature region is identified and a registered type exists, so as to replace the image processing parameters corresponding to the third type in the second memory;and / or a fifth parameter loading module, configured to load, from the first memory to the second memory, image processing parameters corresponding to a type with the highest priority to replace image processing parameters corresponding to the third type in the second memory, when the second feature region is not identified and the registered type does not exist; and / or the image acquisition component is an infrared image acquisition component; and / or the type of the biological feature includes at least two of the following types: palm, face, back of hand, eye, ear, finger.

11. An identity authentication module, characterized by, include: An image acquisition component is used to acquire images and send the acquired images to the processor. A non-volatile first memory is used to store multiple sets of image processing parameters, each set of image processing parameters corresponding to a type of biometric feature; A volatile second memory is used to load one set of image processing parameters from the plurality of image processing parameters; The processor is configured to perform the method according to any one of claims 1-8.

12. The module according to claim 11, characterized in that, The image acquisition component is an infrared image acquisition component.

13. The module of claim 11, wherein The module also includes: An infrared lamp, which is communicatively connected to the image acquisition component, is used to emit infrared light in response to a lighting command sent by the image acquisition component.

14. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method of any one of claims 1-8.

Images