Identity authentication method, device, computer equipment and storage medium

Abstract

The present application discloses an identity authentication method, apparatus, computer equipment, and storage medium, belonging to the field of computer technology. The present application utilizes features stored in a first palmprint feature library for priority palmprint matching, thereby eliminating the need to store all palmprint features in the first palmprint feature library. After the pre-stored palmprint features are loaded into the first palmprint feature library based on the association between pre-stored facial features and pre-stored palmprint features, the palmprint recognition process does not require matching all features in the library one by one. This is equivalent to using facial images to narrow the palmprint matching range, thereby increasing the probability of successfully matching the palmprint features to be detected using the first palmprint feature library. This significantly reduces the computational overhead of the palmprint recognition process and improves the efficiency of identity authentication based on palmprint recognition.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to an identity authentication method, apparatus, computer equipment, and storage medium. Background Art

[0002] With the development of computer technology, when performing certain operations with a high data security level (such as value transfer operations), objects need to authenticate their identities. Currently, in addition to entering a password for identity authentication, identity authentication based on palm print features is also provided.

[0003] During palmprint recognition, the subject reaches out to the camera, which captures an image of the subject's palm, uses it to perform palmprint recognition, and returns the subject's identity authentication result. This palmprint recognition process suffers from high computational overhead and low efficiency due to the large amount of data in the underlying feature library. Summary of the Invention

[0004] The present invention provides an identity authentication method, apparatus, computer device, and storage medium that can reduce the computational overhead of palmprint recognition and improve the efficiency of identity authentication based on palmprint recognition. The technical solution is as follows:

[0005] In one aspect, an identity authentication method is provided, the method comprising:

[0006] In response to an identity authentication operation initiated by a target object, determining pre-stored facial features of the target object based on a captured facial image of the target object;

[0007] adding the pre-stored palmprint feature bound to the pre-stored facial feature to a first palmprint feature library, the first palmprint feature library being used to store features that are prioritized for palmprint matching;

[0008] Extracting the palmprint features to be detected of the target object based on the collected palm image of the target object;

[0009] The palmprint feature to be detected is matched with each palmprint feature stored in the first palmprint feature library, and the identity authentication result of the target object is determined based on the matched palmprint features.

[0010] In one aspect, an identity authentication device is provided, the device comprising:

[0011] a feature determination module, configured to determine, in response to an identity authentication operation initiated by a target object, pre-stored facial features of the target object based on a captured facial image of the target object;

[0012] an adding module for adding the pre-stored palmprint feature bound to the pre-stored facial feature to a first palmprint feature library, wherein the first palmprint feature library is used to store features that are prioritized for palmprint matching;

[0013] An extraction module, configured to extract the palmprint features to be detected of the target object based on the collected palm image of the target object;

[0014] The result determination module is used to match the palmprint feature to be detected with each palmprint feature stored in the first palmprint feature library, and determine the identity authentication result of the target object based on the matched palmprint features.

[0015] In some embodiments, the result determination module includes:

[0016] an acquisition submodule, configured to acquire a palmprint similarity between the palmprint feature to be detected and each palmprint feature stored in the first palmprint feature library;

[0017] a result determination submodule for determining an identity authentication result of the target object based on the matched palmprint feature if a palmprint feature whose palmprint similarity meets a palmprint matching condition exists in the first palmprint feature library, wherein the palmprint matching condition indicates that the palmprint feature to be detected and the palmprint feature belong to the same object;

[0018] The query and determination submodule is used to query palmprint features whose palmprint similarity meets the palmprint matching conditions from the second palmprint feature library if there is no palmprint feature whose palmprint similarity meets the palmprint matching conditions in the first palmprint feature library, and determine the identity authentication result of the target object based on the queried palmprint features. The second palmprint feature library is used to store the full set of features for palmprint matching.

[0019] In some embodiments, the query determination submodule includes:

[0020] a query unit, configured to query, based on the palmprint feature to be detected, candidate palmprint features whose palmprint similarity meets the palmprint matching condition from the second palmprint feature database;

[0021] A result returning unit, configured to, if there is only one candidate palmprint feature, set the identity authentication result as passed palmprint recognition and return the identity identifier bound to the candidate palmprint feature;

[0022] A result determination unit is configured to determine the identity authentication result of the target object based on the candidate facial features bound to the candidate palmprint features if there is more than one candidate palmprint feature.

[0023] In some embodiments, the result determination unit includes:

[0024] a determination subunit, configured to determine, if there is more than one candidate palmprint feature, a palmprint similarity difference between the candidate palmprint feature with the greatest palmprint similarity and each of the remaining candidate palmprint features;

[0025] A result returning subunit is configured to set the identity authentication result as passed palmprint recognition if the palmprint similarity difference values of all palmprints are not less than a difference threshold, and return the identity identifier bound to the candidate palmprint feature with the greatest palmprint similarity;

[0026] The result determination subunit is configured to determine the identity authentication result of the target object based on the candidate facial features bound to the candidate palmprint features if the similarity difference of any palmprint is less than the difference threshold.

[0027] In some embodiments, the result determination subunit includes:

[0028] The first determining sub-subunit is configured to determine the candidate palmprint feature with the greatest palmprint similarity and the candidate palmprint feature with a palmprint similarity difference less than the difference threshold as the target candidate palmprint feature;

[0029] an acquisition sub-subunit, configured to acquire facial similarity between the candidate facial feature bound to each target candidate palmprint feature and the pre-stored facial feature;

[0030] a rejection sub-subunit, configured to reject target candidate palmprint features whose facial similarity does not meet a facial matching condition, to obtain retained candidate palmprint features, wherein the facial matching condition indicates that the candidate facial features and the pre-stored facial features belong to the same object;

[0031] The second determining sub-subunit is configured to determine an identity authentication result of the target object based on the retained candidate palmprint features.

[0032] In some embodiments, the second determining sub-subunit is configured to:

[0033] If there is only one candidate palmprint feature left, the identity authentication result is set to pass palmprint recognition, and the identity identifier bound to the candidate palmprint feature is returned;

[0034] If there are multiple candidate palmprint features left, the identity authentication result is set to fail palmprint recognition.

[0035] In some embodiments, the feature determination module is configured to:

[0036] Extracting facial features to be detected of the target object based on the collected facial image of the target object;

[0037] The facial features to be detected are matched with various facial features stored in a facial feature library, and the matched facial features are determined as pre-stored facial features of the target object.

[0038] In some embodiments, the result determination module is further configured to:

[0039] If no matching facial features are found in the facial feature library, matching the palmprint features to be detected with each palmprint feature stored in a second palmprint feature library, and determining an identity authentication result of the target object based on the matched palmprint features, wherein the second palmprint feature library is used to store all palmprint matching features;

[0040] The facial features to be detected are used as pre-stored facial features of the target object, and the pre-stored facial features are bound to the palm print features matched in the second palm print feature library.

[0041] In some embodiments, the apparatus further comprises:

[0042] The prompting module is used to prompt the target object to retake the facial image or palm image if the collected facial image or palm image of the target object does not meet the image detection conditions.

[0043] In some embodiments, the apparatus further comprises:

[0044] The synchronization module is configured to synchronize the palmprint features stored in the first palmprint feature library in response to a data synchronization instruction of the first palmprint feature library.

[0045] In some embodiments, the apparatus further comprises:

[0046] The acquisition module is used to acquire a facial image of the target object in response to the target object performing an identity authentication operation; and to acquire a palm image of the target object in response to the target object performing a palm extension operation.

[0047] On the one hand, a computer device is provided, which includes one or more processors and one or more memories, wherein at least one computer program is stored in the one or more memories, and the at least one computer program is loaded and executed by the one or more processors to implement the identity authentication method as described above.

[0048] On the one hand, a computer-readable storage medium is provided, in which at least one computer program is stored. The at least one computer program is loaded and executed by a processor to implement the above-mentioned identity authentication method.

[0049] In one aspect, a computer program product is provided, comprising one or more computer programs stored in a computer-readable storage medium. One or more processors of a computer device can read the one or more computer programs from the computer-readable storage medium and execute the one or more computer programs, thereby enabling the computer device to perform the aforementioned identity authentication method.

[0050] The beneficial effects of the technical solutions provided in the embodiments of the present application include at least:

[0051] By utilizing the features stored in the first palmprint feature library for priority palmprint matching, it is not necessary to store all palmprint features in the first palmprint feature library. After the pre-stored palmprint features are loaded into the first palmprint feature library through the association between the pre-stored facial features and the pre-stored palmprint features, since it is not necessary to match all the features in the library one by one during the palmprint recognition process, it is equivalent to using the facial image to narrow the palmprint matching range, thereby improving the probability of successfully matching the palmprint features to be detected using the first palmprint feature library, greatly reducing the computational overhead of the palmprint recognition process, and improving the efficiency of identity authentication based on palmprint recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0053] Figure 1 This is a schematic diagram of an implementation environment of an identity authentication method provided in an embodiment of the present application;

[0054] Figure 2 This is a flow chart of an identity authentication method provided in an embodiment of the present application;

[0055] Figure 3 This is a flow chart of an identity authentication method provided in an embodiment of the present application;

[0056] Figure 4 This is a schematic diagram of the principle of an identity authentication method provided in an embodiment of the present application;

[0057] Figure 5 This is a schematic diagram of another identity authentication method provided in an embodiment of the present application;

[0058] Figure 6 This is a flowchart of binding facial features and palm print features provided by an embodiment of the present application;

[0059] Figure 7 This is a flow chart of a method of using a first palmprint feature library for priority comparison provided by an embodiment of the present application;

[0060] Figure 8 This is a flow chart of an embodiment of the present application providing a method of using facial features for auxiliary recognition;

[0061] Figure 9 This is a structural diagram of an identity authentication device provided in an embodiment of the present application;

[0062] Figure 10 It is a structural diagram of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0063] In order to make the objectives, technical solutions and advantages of this application clearer, the implementation methods of this application will be further described in detail below with reference to the accompanying drawings.

[0064] In this application, the terms "first", "second", etc. are used to distinguish identical or similar items with substantially the same effects and functions. It should be understood that there is no logical or temporal dependency between "first", "second", and "nth", nor is there any limitation on the quantity and execution order.

[0065] In this application, the term "at least one" means one or more, and the term "plurality" means two or more. For example, a plurality of palm print features refers to two or more palm print features.

[0066] In this application, the term "including at least one of A or B" refers to the following situations: including only A, including only B, and including both A and B.

[0067] The user-related information (including but not limited to the user's device information, personal information, behavioral information, etc.), data (including but not limited to data used for analysis, stored data, displayed data, etc.), and signals involved in this application, when applied to specific products or technologies using the methods of the embodiments of this application, are all permitted, agreed, authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant information, data, and signals must comply with the relevant laws, regulations, and standards of the relevant countries and regions. For example, the facial images, palm images, facial features, and palm print features involved in this application are all obtained with full authorization.

[0068] The following explains the terms involved in the embodiments of the present application.

[0069] A camera (also known as a computer camera, computer eye, or electronic eye) is a video input device widely used in video conferencing, telemedicine, and real-time observation. Users use the camera to communicate with each other online, using both visual and audio signals. It can also be used in a variety of popular digital imaging and audio / video processing applications.

[0070] Color image: A color image of natural light captured by the camera's color sensor. In "face payment" (a payment method based on facial recognition), it is generally used for: optimizing facial images and comparing and identifying facial features.

[0071] Infrared image: The infrared image captured by the camera's infrared sensor is a pan-infrared light imaging. In "face payment", it is generally used for: liveness detection of the object to be inspected and assisting in comparative recognition of facial features.

[0072] Palmprint features: This refers to the palm image captured by the camera. A pre-trained palm feature extraction model extracts a series of features, ultimately presenting them as a feature vector (i.e., palmprint features) in a multidimensional space. Subsequent comparisons of different palmprint features can be used to determine palmprint similarity and determine whether they belong to the same person.

[0073] Facial features: This refers to the facial image captured by the camera. A pre-trained facial feature extraction model extracts a series of features, ultimately presenting them as a feature vector (i.e., facial features) in a multidimensional space. Subsequent comparisons of different facial features can be used to determine facial similarity and determine whether they are the same person.

[0074] During a palm-swipe payment process, after the target object (such as a user) selects the palm-swipe payment method, the camera of the palm-swipe device is responsible for capturing the palm image of the target object and using it as the basis for determining the identity of the target object. If the identity authentication result is palm print recognition, the subsequent numerical value transfer (such as payment) operation is performed. Although the camera will simultaneously obtain the color palm image and infrared palm image of the target object through the color sensor and infrared sensor respectively, it is convenient to compare and find the matching palm print features in the underlying feature library, and then return the identity of the target object. However, with the expansion of the palm-swipe payment business volume, the amount of data in the underlying feature library is getting larger and larger. At this time, the palm-swipe payment efficiency is low and the accuracy of the palm-swipe payment will also be affected if the entire library is compared and searched. It is impossible to meet the accuracy and real-time requirements required for the electronic payment level.

[0075] In light of this, the present invention proposes a method for palmprint recognition that utilizes the target person's facial features to assist in palmprint recognition. This palmprint recognition solution is applicable to palmprint payment scenarios or other scenarios that require invoking a palmprint payment interface for identity authentication. Even if the underlying feature library data volume is large, it can still improve the efficiency and security of palmprint payment, meet the accuracy and real-time requirements required for electronic payment, and optimize the speed experience of palmprint payment for users.

[0076] The following describes the system architecture of the embodiment of the present application.

[0077] Figure 1 This is a schematic diagram of an implementation environment of an identity authentication method provided in an embodiment of the present application. Figure 1 The implementation environment includes: a palm-brush device 120 and a server 140. The palm-brush device 120 and the server 140 can be directly or indirectly connected via wired or wireless communication, which is not limited in this application. The palm-brush device 120 and the server 140 are both exemplary descriptions of computer devices.

[0078] The palm-swipe device 120 is used to support palm-swipe payment services. For example, the palm-swipe device 120 includes a camera 122 and a host 124. The camera 122 is used to capture facial and palm images and send the captured facial and palm images to the host 124. If the host 124 has a first palmprint feature library cached locally, the identity authentication method of the embodiment of the present application can be executed locally to complete palmprint recognition and, if palmprint recognition succeeds, to proceed with the subsequent payment process. Alternatively, the host 124 compresses, encrypts, and encapsulates the facial data and palm image to generate a palmprint payment request, which is then sent to the server 140 to request the server 140 to execute the identity authentication method of the embodiment of the present application, complete palmprint recognition, and, if palmprint recognition succeeds, to proceed with the subsequent payment process.

[0079] Optionally, the palm-swiping device 120 is a palm-swiping payment terminal, a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, etc., but is not limited thereto.

[0080] In some embodiments, the camera 122 may be a 3D camera having functions such as facial recognition, gesture recognition, human skeleton recognition, three-dimensional measurement, environmental perception, or three-dimensional map reconstruction. The 3D camera can detect the distance information between each pixel in the captured image and the camera to obtain a depth map, thereby determining whether the object corresponding to the currently captured image is alive, thereby avoiding security risks caused by attackers using other people's photos for identity authentication.

[0081] Server 140 may include at least one of a single server, multiple servers, a cloud computing platform, or a virtualization center. Server 140 is used to provide background services for applications running on the palm-swiping device 120. The application can provide users with palm-swiping payment services, allowing users to transfer values based on the palm-swiping device 120. Optionally, server 140 can perform primary computing tasks, while palm-swiping device 120 can perform secondary computing tasks; alternatively, server 140 can perform secondary computing tasks, while palm-swiping device 120 can perform primary computing tasks; alternatively, server 140 and palm-swiping device 120 can perform collaborative computing tasks using a distributed computing architecture.

[0082] In an exemplary scenario, the palm-swiping device 120 refers to a computer device with an integrated camera that can capture images of the user's palm for payment. For example, the palm-swiping device 120 can be a terminal that provides palm-swiping payment functions, or a payment device provided by a merchant, or an unmanned vending machine. Schematically, after the user triggers the payment option on the palm-swiping device 120 and selects the palm-swiping payment method, the palm-swiping device 120 will turn on the camera to capture the user's facial and palm images after obtaining full authorization from the user. Optionally, after the palm-swiping device 120 completes palm print recognition locally, if the palm print recognition is successful, it interacts with the server 140 to transfer the value; alternatively, the palm-swiping device 120 requests the server 140 to perform palm print recognition in the cloud. If the palm print recognition is successful, the server 140 transfers the value.

[0083] Optionally, server 140 is an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms.

[0084] Optionally, the number of the palm-brushing devices 120 may be more or less. For example, there may be only one palm-brushing device 120, or there may be dozens, hundreds, or even more of the palm-brushing devices 120. This embodiment of the application does not limit the number or type of the palm-brushing devices 120.

[0085] The following describes the basic process of the identity authentication method according to the embodiment of the present application.

[0086] Figure 2 This is a flow chart of an identity authentication method provided by an embodiment of the present application. Figure 2This embodiment is executed by a computer device, for example, the computer device is the palm-swipe device 120 in the above implementation environment or other terminals with palm-swipe payment function, and includes the following steps:

[0087] 201. In response to an identity authentication operation initiated by a target object, a computer device determines pre-stored facial features of the target object based on a collected facial image of the target object.

[0088] The target object refers to any object that accesses the identity authentication service. The target object can access the identity authentication service directly or indirectly. The identity authentication service can be accessed separately, such as when the target object registers for the identity authentication service for the first time. The identity authentication service can also be a link in other services. For example, the target object calls the identity authentication service to ensure security during the value transfer operation. For example, the target object calls the identity authentication service to ensure security during the modification of attribute information.

[0089] In some embodiments, the target object initiates an identity authentication operation on a client of a computer device, causing the computer device to respond to the identity authentication operation and apply for camera permission. After the target object fully authorizes and agrees, the camera is turned on to capture the target object's facial image, and based on the captured facial image, the target object's facial features to be detected are extracted, the facial image to be detected is matched with each facial feature stored in a facial feature library, and the matched facial features are determined as the pre-stored facial features of the target object. Alternatively, if no facial features are matched in the facial feature library, it means that the target object is using the identity authentication service for the first time, and therefore the registration process for the identity authentication service can be started, the facial features to be detected are stored as the pre-stored facial features of the target object in the facial feature library, and the pre-stored facial features are bound to the palm print features matched in step 204, that is, an association relationship is established between the pre-stored facial features and the matched palm print features. The embodiment of the present application does not specifically limit the method for obtaining the pre-stored facial features.

[0090] 202. The computer device adds the pre-stored palmprint feature bound to the pre-stored facial feature to a first palmprint feature library, where the first palmprint feature library is used to store features that are prioritized for palmprint matching.

[0091] In some embodiments, the computer device queries the association between facial features and palm print features, finds the pre-stored palm print features bound to the pre-stored facial features, and then adds the pre-stored palm print features to a first palm print feature library. The first palm print feature library can be a local database of the computer device or a cloud database in a cloud server.

[0092] Among them, the first palmprint feature library is relative to the second palmprint feature library. The second palmprint feature library stores the full amount of features for palmprint matching. The first palmprint feature library is a priority feature library with a smaller data volume than the second palmprint feature library. It is used to store features for priority palmprint matching. Optionally, it is used to store the palmprint features of objects that have accessed the identity authentication service in the recent period, or to store the palmprint features of objects that have accessed the identity authentication service more frequently in the recent period, or to store the palmprint features of objects that have accessed the identity authentication service locally in the recent period.

[0093] In one example, the first palmprint feature library is a local database of a computer device. In this way, when palmprint matching is performed through step 203, there is no need to perform palmprint matching with the second palmprint feature library in the cloud. This can avoid the time-consuming request for identity authentication from the cloud, save the communication overhead of interacting with the cloud, improve the efficiency of identity authentication, and thus improve the efficiency of value transfer. If the first palmprint feature library fails to match successfully, the second palmprint feature library in the cloud is requested to perform palmprint matching based on the full set of features. This can maximize the success rate of identity authentication and thus improve the success rate of value transfer.

[0094] 203. The computer device extracts the palmprint features to be detected of the target object based on the collected palm image of the target object.

[0095] In some embodiments, after applying for camera access and obtaining full authorization and consent from the target subject, the computer device turns on the camera to continue capturing palm images of the target subject, and extracts the target subject's palmprint features to be tested based on the captured palm images. Optionally, the camera captures a color palm image and an infrared palm image, which are then input into a pre-trained palmprint feature extraction model, and the palmprint features to be tested are extracted using the palmprint feature extraction model.

[0096] 204. The computer device matches the palmprint feature to be detected with each palmprint feature stored in the first palmprint feature library, and determines the identity authentication result of the target object based on the matched palmprint features.

[0097] In some embodiments, the computer device determines whether the palmprint feature to be detected matches each palmprint feature stored in the first palmprint feature library one by one, and determines the target object's identity authentication result based on the matched palmprint features. The identity authentication result indicates whether the target object passed the palmprint recognition. Optionally, if the target object passed the palmprint recognition, the identity authentication result also carries the target object's bound identity identifier. If the target object did not pass the palmprint recognition, the target object may be prompted to initiate the identity authentication operation again. The palmprint recognition process will be described in detail in the next embodiment and will not be repeated here.

[0098] It should be noted that the identity authentication process provided in the above steps 201-204 can be used as a complete process for a separate identity authentication business, or as an authentication link in the process of other businesses that need to call the identity authentication interface. This embodiment of the present application does not specifically limit this.

[0099] The method provided in the embodiment of the present application utilizes the features stored in the first palmprint feature library for priority palmprint matching, so that the full set of palmprint features does not need to be stored in the first palmprint feature library. After the pre-stored palmprint features are loaded into the first palmprint feature library through the association between the pre-stored facial features and the pre-stored palmprint features, since the palmprint recognition process does not require matching all the features in the library one by one, it is equivalent to using the facial image to narrow the palmprint matching range, thereby increasing the probability of successfully matching the palmprint features to be tested using the first palmprint feature library, greatly reducing the computational overhead of the palmprint recognition process, and improving the efficiency of identity authentication based on palmprint recognition. Furthermore, since the pre-stored facial features are used to lock the pre-stored palmprint features, the matching palmprint features found in the first palmprint feature library are more likely to be the pre-stored palmprint features, thereby also improving the success rate of identity authentication based on palmprint recognition.

[0100] All the above optional technical solutions can be arbitrarily combined to form optional embodiments of the present disclosure, and will not be described in detail here.

[0101] The previous embodiment briefly introduced the basic process of the identity authentication method. In this embodiment, the detailed process of the identity authentication method of this embodiment will be described using a computer device as a palm-swipe device. In one example, the palm-swipe device can be a payment terminal configured by a merchant to facilitate palm-swipe payment, or an unmanned vending machine with an integrated palm-swipe payment function, or other computer devices with palmprint recognition capabilities.

[0102] Figure 3 This is a flow chart of an identity authentication method provided by an embodiment of the present application. Figure 3 This embodiment is executed by a palm brushing device and includes the following steps:

[0103] 301. The palm-scanning device collects a facial image of the target object in response to an identity authentication operation initiated by the target object.

[0104] In some embodiments, the target subject can initiate an identity authentication operation on the palm-swiping device's client. The palm-swiping device then performs the identity authentication operation in response to the target subject, applies for camera access, and, after receiving full authorization and consent from the target subject, opens the camera to capture the target subject's facial image. Alternatively, the palm-swiping device captures a facial image using its camera, or captures a facial video stream using its camera and selects one frame from multiple frames in the facial video stream as the captured facial image. The facial image capture method is not specifically limited herein.

[0105] In some embodiments, the palm-scanning device determines whether the captured facial image of the target subject meets the image detection conditions. If the captured facial image of the target subject does not meet the image detection conditions, the target subject is prompted to retake the facial image. If the captured facial image of the target subject meets the image detection conditions, the process proceeds to step 302. The image detection conditions indicate the quality standards of an image that can be used for identity authentication. For example, the image detection conditions include, but are not limited to, clarity, resolution, and the proportion of the target subject's recognized facial area in the facial image. The specific content of the image detection conditions is not specifically limited herein.

[0106] In the above process, by determining whether the collected facial image meets the preset image detection conditions, the image quality of the facial image used to determine the pre-stored facial features can be guaranteed, which helps improve the accuracy of obtaining the bound pre-stored palmprint features, thereby indirectly improving the success rate of palmprint recognition and helping to improve the palmprint recognition experience. It should be noted that the above step of determining the image detection conditions is an optional step, and it is also possible not to determine whether the collected facial image meets the image detection conditions, which can simplify the palmprint recognition process.

[0107] In some embodiments, taking the palm-swipe payment scenario as an example, the merchant is equipped with a palm-swipe device in the venue. After the cashier enters the value to be transferred, or the target object enters the value to be transferred when self-checking out, the value transfer process is started. In the value transfer process, the identity of the target object needs to be verified first. Only after the identity verification of the target object is passed, the value to be transferred will be deducted from the target object's account balance, and the value to be transferred will be added to the merchant's account balance. Optionally, when verifying the identity of the target object, a variety of optional identity authentication methods are provided. When the target object selects the palm-swipe recognition method (such as the palm-swipe payment method), the palm-swipe recognition process is started. In the palm-swipe recognition process, the camera permission is first requested (such as displaying an authorization pop-up window, providing an authorization button, etc.). After the target object has fully authorized and agreed, the camera is turned on to collect the target object's facial image.

[0108] 302. The palm-scanning device determines pre-stored facial features of the target object based on the collected facial image of the target object.

[0109] In some embodiments, the palm-scanning device extracts the target object's facial features to be tested based on the facial image of the target object captured in step 301. The facial features to be tested are then matched with each facial feature stored in a facial feature library, and the matched facial features are determined as the pre-stored facial features of the target object. In other words, the above process provides a possible implementation for determining the pre-stored facial features of the target object. After extracting the facial features to be tested using the facial image, the facial features to be tested are compared one by one with each facial feature stored in the facial feature library. For example, the facial similarity between the facial features to be tested and each facial feature stored in the facial feature library is calculated one by one. Next, a determination is made as to whether the facial similarity is greater than a preset facial similarity threshold. If the facial similarity is greater than the preset facial similarity threshold, it indicates that the facial feature stored in the facial feature library and the facial feature to be tested match. If the facial similarity is less than the preset facial similarity threshold, it indicates that the facial feature stored in the facial feature library and the facial feature to be tested do not match. The facial similarity between the next facial feature and the facial feature to be tested is then determined. The above process is repeated until all facial features stored in the facial feature library have been traversed. At this time, if no matching facial features are detected, it means that the target object is a new customer who uses the identity authentication service for the first time, and the target object can be prompted to register for the identity authentication service; if only one matching facial feature is detected, the matching facial feature is directly used as the pre-stored facial feature of the target object; if more than one matching facial feature is detected, the facial feature with the highest facial similarity can be determined as the pre-stored facial feature of the target object.

[0110] The facial similarity between each pair of facial features is used to characterize whether the two facial features are similar (i.e., the degree of similarity), thereby reflecting the likelihood that the two facial features belong to the same object (i.e., the matching probability). A larger facial similarity value indicates a higher degree of similarity between the facial features, and a higher matching probability that the two facial features belong to the same object. Conversely, a smaller facial similarity value indicates a lower degree of similarity between the facial features, and a lower matching probability that the two facial features belong to the same object. The facial similarity can be the cosine similarity between the two facial features, or the inverse of the Euclidean distance between the two facial features, or can be calculated using other calculation methods. The method for measuring facial similarity is not specifically limited herein.

[0111] In the above process, facial features to be detected are extracted from the facial image, and then the facial features to be detected are matched with the facial features stored in the facial feature library to find the pre-stored facial features of the target object when it was registered. In this way, in step 303 below, based on the correlation between facial features and palm print features, the pre-stored facial features are used to pre-lock a pre-stored palm print feature with a higher probability, thereby narrowing the matching range of palm print recognition and improving the efficiency and success rate of palm print recognition. It should be noted that the facial features stored in the facial feature library here are used to assist in palm print recognition and do not need to reach the feature refinement level of facial recognition services. Therefore, the overhead of feature extraction and matching in this step is relatively controllable and can be compressed within the real-time requirements of the payment level.

[0112] In some embodiments, when extracting facial features to be detected using facial images, the palm-scanning device may locally store a pre-trained facial feature extraction model. After turning on the camera, the color sensor captures a color facial image and the infrared sensor captures an infrared facial image. The color and infrared facial images are then input into the locally stored facial feature extraction model. The facial feature extraction model then performs feature extraction on the color and infrared facial images to obtain the facial features to be detected. The facial feature extraction model is used to extract features of the facial region of the object identified in the input image.

[0113] 303. The palm scanning device adds the pre-stored palm print feature bound to the pre-stored facial feature to a first palm print feature library, where the first palm print feature library is used to store features that are prioritized for palm print matching.

[0114] In some embodiments, since each object will bind its own facial features and palm print features provided during registration (i.e., establish an association relationship) when registering for the identity authentication service, the palm-scanning device queries the association relationship between the facial features and the palm print features based on the pre-stored facial features of the target object determined in step 302, and can find the pre-stored palm print features bound to the pre-stored facial features. Then, the pre-stored palm print features are added to the first palm print feature library. The first palm print feature library can be a local database of the palm-scanning device or a cloud database in a cloud server.

[0115] In some embodiments, the association between facial features and palm print features can be implemented as an association table, in which the association between the facial features and palm print features of each object registered for the identity authentication service is stored in a Key-Value data structure. In one example, for each object registered for the identity authentication service, the palm print features entered by the object during registration and the facial features bound to the palm print features are recorded in a Key-Value data structure. For example, the facial feature ID (Identification) of the facial features of the object is used as the key name Key, and the palm print feature ID of the palm print features of the object is used as the key value Value.

[0116] In this case, the facial feature ID of the pre-stored facial feature determined in step 302 is used as an index to query whether the Key-Value data structure can be hit in the association relationship table. If any Key-Value data structure can be hit, the palm print feature ID in the Value is taken out, and the bound pre-stored palm print feature can be reversely searched from the second palm print feature library based on the palm print feature ID; if no Key-Value data structure is hit, it means that the target object has not registered for the identity authentication service, or has not bound its facial features after registering for the identity authentication service, and the target object can be prompted to enter the facial features.

[0117] In some embodiments, the first palmprint feature library is a local database of the palm-swiping device. In this way, when palmprint matching is performed through steps 306-308, it is not necessary to perform palmprint matching with the second palmprint feature library in the cloud every time. This can avoid the time-consuming request for identity authentication from the cloud, save the communication overhead of interacting with the cloud, improve the efficiency of identity authentication, and thus improve the efficiency of value transfer. Only when the first palmprint feature library fails to match successfully, will the second palmprint feature library in the cloud (or local) be requested to perform palmprint matching based on the full set of features. This can maximize the success rate of identity authentication and thus improve the success rate of value transfer.

[0118] In the embodiment of the present application, the first palmprint feature library is an example of a local database of the palm-scanning device. The full features of the second palmprint feature library are maintained in the cloud, but the palm-scanning device can load a part of the sub-library of the second palmprint feature library that is adapted to the geographical location to the local area according to the geographical location. In this way, when the first palmprint feature library fails to match, the palmprint matching can be performed first from the sub-library of the second palmprint feature library loaded locally. If the match is successful, the identity authentication result can be obtained. If the match fails again, the cloud is requested to use the full second palmprint feature library for palmprint matching. This can ensure that palmprint recognition is completed locally as much as possible, improve the palmprint recognition speed and identity authentication efficiency, and in this case, the palm-scanning device does not need to maintain the full second palmprint feature library locally, thereby greatly saving the storage overhead of the palm-scanning device, and does not require the palm-scanning device to have a large storage capacity, saving the cost of the palm-scanning device.

[0119] In some embodiments, steps 302-303 provide a method for updating the first palmprint feature library when a facial feature matching the facial feature to be detected has been found in the facial feature library. Specifically, by extracting the facial feature to be detected using a facial image, and then matching the facial feature to be detected with the facial features stored in the facial feature library, the pre-stored facial features of the target object at the time of registration are found. This allows the pre-stored facial features to be used to pre-lock a pre-stored palmprint feature with a higher probability based on the correlation between the facial features and the palmprint features. The pre-stored palmprint feature is then stored in the first palmprint feature library, giving priority to using the first palmprint feature library for palmprint matching, eliminating the need to compare all features in the second palmprint feature library one by one. This reduces the matching range of palmprint recognition and improves the efficiency and success rate of palmprint recognition.

[0120] In other embodiments, if no matching facial features are found in the facial feature library, then after obtaining full authorization and consent from the target object, the following steps 304-305 can be executed to obtain the target object's palm print features to be tested, and then the palm print features to be tested are directly matched with the various palm print features stored in the second palm print feature library, and the identity authentication result of the target object is determined based on the matched palm print features, wherein the second palm print feature library is used to store the full set of features of the palm print match. At this time, even if the target object has not established an association relationship between facial features and palm print features, the second palm print feature library can be directly used to perform palm print matching, which ensures the success rate of the identity authentication business and avoids affecting the normal execution of the identity authentication business. In addition, the facial features to be detected can be directly used as the pre-stored facial features of the target object, and the pre-stored facial features are bound to the palm print features matched in the second palm print feature library. At this time, an association relationship between facial features and palm print features is established for the target object. The next time the target object performs identity authentication, it can follow the process of this solution and give priority to using the first palm print feature library for palm print recognition, thereby improving the identity authentication efficiency of the target object when performing identity authentication next time.

[0121] The following is an explanation of a possible binding process of facial features and palm print features. In the case that the target object has not established an association relationship between facial features and palm print features, a facial image is collected through step 301, and facial features of the facial image are extracted as facial features to be detected of the target object. The facial features to be detected are used as pre-stored facial features of the target object, a facial feature ID is assigned to the pre-stored facial features of the target object, and the pre-stored facial features of the target object are stored in a facial feature library. Similarly, the palm print features to be detected of the target object are obtained through steps 304-305. At this time, if the target object is not using the identity authentication service for the first time and has not yet bound facial features to the palm print features, then step 306 can be used to authenticate the target object. -308 A pre-stored palmprint feature matching the palmprint feature to be checked is found, and the palmprint feature ID of the pre-stored palmprint feature is read. If the target object is using the identity authentication service for the first time, the pre-stored palmprint feature cannot be found. The target object can be directly prompted to set the palmprint feature to be checked as the pre-stored palmprint feature, and a palmprint feature ID is assigned to the pre-stored palmprint feature of the target object. The pre-stored palmprint feature is then stored in the second palmprint feature database. Next, a key-value data structure is constructed with the facial feature ID of the pre-stored facial feature of the target object as the key and the palmprint feature ID of the pre-stored palmprint feature of the target object as the value. The constructed key-value data structure is inserted into the association relationship table.

[0122] In some embodiments, the first palmprint feature library is a local database of the palm-swiping device, but data synchronization can be periodically enabled between different palm-swiping devices to ensure timely updating of the first palmprint feature library and further improve the accuracy of the first palmprint feature library. At this time, data synchronization can be assisted by a server in the cloud, that is, the server periodically issues data synchronization instructions for the first palmprint feature library, or other palm-swiping devices can initiate D2D (Device to Device) communication to transmit the data synchronization instructions for the first palmprint feature library. The embodiment of the present application does not specifically limit the data synchronization method. After receiving the data synchronization instruction for the first palmprint feature library, the palm-swiping device responds to the data synchronization instruction for the first palmprint feature library and synchronizes the palmprint features stored in the first palmprint feature library.

[0123] In the above process, by synchronizing the first palmprint feature library with data (which can be periodic synchronization or data synchronization triggered by a data synchronization instruction), the first palmprint feature library is ensured to be updated in a timely manner, thereby further improving the accuracy of the first palmprint feature library. In the case of periodic synchronization, there is no need to frequently transmit synchronization data, so the communication overhead is relatively low. In the case of instruction-triggered synchronization, each time a new palmprint feature is added, synchronization can be achieved in a timely manner. Therefore, the synchronization of the first palmprint feature library is more real-time and can be achieved in a timely manner.

[0124] In one example, the incremental feature data of the first palmprint feature library within a certain period of time is synchronized in an incremental synchronization manner. For example, each palm-swiping device will upload to the server the incremental feature data from the time of the last data synchronization (which can be implemented based on the checkpoint CheckPoint method) to the current time at intervals of an upload cycle. In this way, the incremental data of the first palmprint feature library local to the palm-swiping device compared to the last data synchronization can be reported to the server in a timely manner. Then, at intervals of a statistical cycle, the server summarizes all the incremental feature data reported by each palm-swiping device within the statistical cycle, and issues a data synchronization instruction for the first palmprint feature library to each palm-swiping device, so that the palm-swiping device responds to the data synchronization instruction and pulls all the incremental feature data summarized within the statistical cycle from the server. The length of the upload cycle and the length of the statistical cycle can be the same or different, and are not specifically limited here.

[0125] 304. In response to the target object performing a palm extension operation, the palm-swiping device collects a palm image of the target object.

[0126] In some embodiments, the target object can perform a palm-extending operation facing the camera of the palm-swiping device, and the palm-swiping device responds to the target object performing the palm-extending operation and collects the palm image of the target object through the camera. Optionally, the palm-swiping device captures a palm image through the camera, or the palm-swiping device collects a palm video stream through the camera and selects one frame from multiple video frames in the palm video stream as the collected palm image. The method of collecting the palm image is not specifically limited here. It should be noted that the camera permission can be applied for again here, and the full authorization and consent of the target object can be obtained. Alternatively, in order to avoid frequently displaying the authorization pop-up window to disturb the target object, the authorization to collect the facial image and the palm image can be requested only when the camera permission is applied for the first time in step 301. There is no specific limitation here. In the case of applying for the camera permission only once, the facial video stream in step 301 and the palm video stream in step 304 can be a continuously shot video stream or two separately shot video streams.

[0127] In some embodiments, the palm scanning device determines whether the captured palm image of the target object meets the image detection conditions. If the captured palm image of the target object does not meet the image detection conditions, the target object is prompted to retake the palm image. If the captured palm image of the target object meets the image detection conditions, the process proceeds to step 305. The image detection conditions indicate the quality standards of the image that can be used for identity authentication. For example, the image detection conditions include but are not limited to: clarity, resolution, the proportion of the palm area of the target object recognized in the palm image, etc. The specific content of the image detection conditions is not specifically limited here.

[0128] In the above process, by determining whether the collected palm image meets the preset image detection conditions, the image quality of the palm image used to extract the palmprint features to be detected can be guaranteed, which helps to improve the feature quality and feature accuracy of the palmprint features to be detected, thereby indirectly improving the success rate of palmprint recognition and helping to improve the palmprint recognition experience. It should be noted that the above step of determining the image detection conditions is an optional step, and it is also possible not to determine whether the collected palm image meets the image detection conditions, which can simplify the palmprint recognition process.

[0129] In some embodiments, taking the palm swiping payment scenario as an example, after requesting camera permission in the palm swiping recognition process (such as displaying an authorization pop-up window, providing an authorization button, etc.), the camera is turned on to collect a video stream after the target object is authorized, and a frame of facial image and a frame of palm image are sampled from the video stream respectively.

[0130] 305. The palm scanning device extracts the palmprint features to be detected of the target object based on the collected palm image of the target object.

[0131] In some embodiments, the palm scanning device extracts the target object's palm print features to be tested based on the target object's palm image captured in step 304. Optionally, the palm scanning device may locally store a pre-trained palm print feature extraction model. After turning on the camera, the palm scanning device captures a color palm image via the color sensor and an infrared palm image via the infrared sensor. The color palm image and the infrared palm image are then input into the locally stored palm print feature extraction model. The palm print feature extraction model then performs feature extraction on the color palm image and the infrared palm image to obtain the target object's palm print features. The palm print feature extraction model is used to extract features of the palm area of the object identified in the input image.

[0132] 306. The palm scanning device obtains the palm print similarity between the palm print feature to be detected and each palm print feature stored in the first palm print feature library.

[0133] In some embodiments, after the palm scanning device extracts the palm print feature to be detected using the palm image, it compares the palm print feature to be detected with each palm print feature stored in the first palm print feature library one by one. For example, the palm print similarity between the palm print feature to be detected and each palm print feature stored in the first palm print feature library is calculated one by one. The palm print similarity between each two palm print features is used to indicate whether the two palm print features are similar (i.e., the degree of similarity), thereby reflecting the possibility that the two palm print features belong to the same object (i.e., the matching probability). The larger the palm print similarity value, the higher the degree of similarity of the palm print features, which means the matching probability that the two palm print features belong to the same object is greater. Conversely, the smaller the palm print similarity value, the lower the degree of similarity of the palm print features, which means the matching probability that the two palm print features belong to the same object is smaller. The palm print similarity can be the cosine similarity between the two palm print features, or can be the inverse of the Euclidean distance between the two palm print features, or can be set to other calculation methods. The palm print similarity measurement method is not specifically limited here.

[0134] 307. If a palm print feature exists in the first palm print feature database whose palm print similarity meets the palm print matching condition, the palm scanning device determines the identity authentication result of the target object based on the matched palm print feature, and the palm print matching condition indicates that the palm print feature to be detected and the palm print feature belong to the same object.

[0135] In some embodiments, the palmprint matching condition is that the palmprint similarity is greater than a preset palmprint similarity threshold. In this case, each palmprint feature stored in the first palmprint feature library is traversed, and the palmprint similarity between the palmprint feature and the palmprint feature to be tested extracted in step 305 is calculated. Then, it is determined whether the palmprint similarity is greater than the palmprint similarity threshold. If it is greater than the preset palmprint similarity threshold, it means that the palmprint feature stored in the first palmprint feature library and the palmprint feature to be tested match, and it is determined that the palmprint feature meets the palmprint matching condition. If it is not greater than the palmprint similarity threshold, it means that the palmprint feature stored in the first palmprint feature library and the palmprint feature to be tested do not match, and the palmprint similarity between the next palmprint feature and the palmprint feature to be tested is determined.

[0136] After the first palmprint feature library is traversed through the entire library, if no palmprint feature that meets the palmprint matching conditions is found, the process proceeds to step 308. If a palmprint feature that meets the palmprint matching conditions is found, all palmprint features that meet the palmprint matching conditions will be obtained. If there is only one palmprint feature that meets the palmprint matching conditions in the first palmprint feature library, the identity authentication result can be directly set to pass palmprint recognition, and the identity identifier bound to this palmprint feature is returned. If there are more than one palmprint features that meet the palmprint matching conditions in the first palmprint feature library, the following processing methods can be used: 1) The identity authentication result is set to pass palmprint recognition, and the identity identifier bound to this palmprint feature is returned. The palmprint recognition result is set to fail, and the target object is prompted to re-initiate palmprint recognition or switch the identity authentication mode. 2) Among the palmprint features that meet the palmprint matching conditions in the first palmprint feature library, it is determined whether the difference between the palmprint feature with the highest palmprint similarity and the palmprint feature with the second highest palmprint similarity is less than a preset difference threshold. If it is less than the difference threshold, the identity authentication result is set to fail palmprint recognition, and the target object is prompted to re-initiate palmprint recognition or switch the identity authentication mode. If it is not less than the difference threshold, the identity authentication result is set to pass palmprint recognition, and the identity identifier bound to the palmprint feature with the highest palmprint similarity is returned.

[0137] In the above situation, if there is only one matched palmprint feature in the first palmprint feature library, the performance overhead of the full library matching shall not exceed the feature capacity of the first palmprint feature library. In this way, there is no need to access the second palmprint feature library. Since the second palmprint feature library is usually deployed in the cloud, and the palm-swiping device only loads a part of the sub-library locally, there is no need to interact with the cloud server. This avoids communication delays and only requires considering the matching time. This is especially obvious in large-scale scenarios at the million level, and can greatly improve the efficiency and success rate of palmprint recognition.

[0138] In the above situation, if there is more than one matched palmprint feature in the first palmprint feature library, in method 1), since the first palmprint feature library is the object that has accessed the identity authentication service in the recent period, the palmprint features are reversely checked through their facial features to form a priority comparison feature library. At this time, if there are multiple matched palmprint features, it means that the quality of the palm image may be not high. In order to ensure the security of the identity authentication service, it directly returns that the palmprint recognition fails, prompting the target object to re-initiate palmprint recognition or switch the identity authentication method, which can fully ensure the security of the service.

[0139] In the above situation, if there is more than one matching palmprint feature in the first palmprint feature library, in method 2), since the first palmprint feature library is the object that has accessed the identity authentication service in the recent period, the palmprint features are reversely checked through their facial features to form a priority comparison feature library. At this time, if there are multiple matching palmprint features, blindly judging that the palmprint recognition is not passed may reduce the success rate of palmprint recognition and bring damage to the recognition experience of the target object. Therefore, it is further judged whether the difference between the first two palmprint features with the highest palmprint similarity is less than the preset difference threshold, so that it can be evaluated whether the first two palmprint features are Whether it belongs to "high similarity features"; if it is less than the difference threshold, it means that the first two palmprint features are "high similarity features". At this time, it is impossible to accurately judge which identity the target object belongs to. Palmprint recognition will bring certain risks. Therefore, the identity authentication result is set to fail palmprint recognition, and the target object is prompted to re-initiate palmprint recognition or switch the identity authentication method. If it is not less than the difference threshold, it means that the first two palmprint features are not "high similarity features". In this case, the risk of passing palmprint recognition is relatively low. In this case, the identity authentication result is set to pass palmprint recognition, and the identity bound to the palmprint feature with the highest palmprint similarity is returned.

[0140] 308. If there is no palm print feature whose palm print similarity meets the palm print matching condition in the first palm print feature library, the palm scanning device queries the second palm print feature library for a palm print feature whose palm print similarity meets the palm print matching condition, and determines the identity authentication result of the target object based on the queried palm print feature. The second palm print feature library is used to store all the features of the palm print match.

[0141] In some embodiments, after the first palmprint feature library is traversed in its entirety in step 307, if no palmprint features meeting the palmprint matching conditions are found, it means that no identity identifier matching the target object is found in the priority comparison feature library, and then the matching range needs to be expanded. At this time, the palm-swiping device can further query the second palmprint feature library to ensure the success rate of palmprint recognition.

[0142] In some embodiments, the second palmprint feature library can be a database maintained in the cloud by a server providing palmprint recognition services. In this case, the palm-scanning device can call the palmprint recognition interface provided by the server and send an RPC (Remote Procedure Call) request through the palmprint recognition interface. The RPC request is a method of implementing a palmprint recognition request. The RPC request carries at least the target object's palmprint features to be tested. The RPC request is used to request the server to perform palmprint recognition on the palmprint features to be tested and return the target object's identity authentication result. Of course, the second palmprint feature library can also be a local database of the palm-scanning device. This way, the palmprint recognition process does not require the participation of the requesting server, reducing the communication overhead and communication delay between the palm-scanning device and the server, and improving the efficiency of palmprint recognition.

[0143] In other embodiments, in order to further improve the efficiency of palmprint recognition and reduce the communication overhead and communication delay between the palm-swiping device and the server, the second palmprint feature library on the server side can be divided into multiple sub-libraries according to different geographical locations (this can be used for distributed storage and backup to ensure high service availability). In this case, the palm-swiping device can determine its own geographical location, and then, according to the geographical location, load the sub-library of the second palmprint feature library that is adapted to the geographical location from the server to the local location. In this way, when the first palmprint feature library fails to match, the palmprint matching can be performed from the sub-library of the second palmprint feature library loaded locally. If the match is successful, the identity authentication result can be obtained. If the match fails again, the cloud can be requested to use the full second palmprint feature library for palmprint matching. This can ensure that palmprint recognition is completed locally as much as possible, improve the palmprint recognition speed and identity authentication efficiency, and in this case, the palm-swiping device does not need to maintain the full second palmprint feature library locally, thereby greatly saving the storage overhead of the palm-swiping device, and does not require the palm-swiping device to have a large storage capacity, saving the cost of the palm-swiping device.

[0144] Assume that the second palmprint feature database is a local database of the palm scanning device. Since the palmprint matching process of the second palmprint feature database is the same as that of the sub-database, only the palmprint matching process of the full database is used as an example for explanation. Please refer to the following steps A1 to A3:

[0145] A1. The palm scanning device searches the second palm print feature database for candidate palm print features whose palm print similarity meets the palm print matching condition based on the palm print feature to be detected.

[0146] In some embodiments, the palm scanning device obtains the palm print similarity between the palm print feature to be detected and each palm print feature stored in the second palm print feature library, that is, the palm print similarity between the palm print feature to be detected and each palm print feature stored in the second palm print feature library is calculated one by one. The palm print similarity calculation method is similar to step 306 and will not be repeated here. Next, the palmprint matching condition that the palmprint similarity is greater than a preset palmprint similarity threshold is used as an example for explanation. After traversing each palmprint feature stored in the second palmprint feature library and calculating the palmprint similarity between the palmprint feature and the palmprint feature to be detected, it is determined whether the palmprint similarity is greater than the palmprint similarity threshold. If it is greater than the preset palmprint similarity threshold, it means that the palmprint feature stored in the second palmprint feature library and the palmprint feature to be detected match, and it is determined that the palmprint feature meets the palmprint matching condition, and the palmprint feature is marked as a candidate palmprint feature. If it is not greater than the palmprint similarity threshold, it means that the palmprint feature stored in the second palmprint feature library and the palmprint feature to be detected do not match, and the palmprint similarity between the next palmprint feature and the palmprint feature to be detected is determined.

[0147] After searching the entire second palmprint feature library, if a candidate palmprint feature that meets the palmprint matching criteria is found, all candidate palmprint features that meet the palmprint matching criteria will be obtained. Then, if there is only one candidate palmprint feature that meets the palmprint matching criteria in the second palmprint feature library, the process proceeds to step A2; if there are more than one candidate palmprint feature that meets the palmprint matching criteria in the second palmprint feature library, the process proceeds to step A3. In addition, a scenario is also considered where, after searching the entire second palmprint feature library, if no candidate palmprint feature that meets the palmprint matching criteria is found, the identity authentication result is set as failed palmprint recognition, prompting the target subject to re-initiate palmprint recognition or switch the identity authentication method.

[0148] A2. If there is only one candidate palmprint feature, the palm scanning device sets the identity authentication result as passed palmprint recognition and returns the identity identifier bound to the candidate palmprint feature.

[0149] In some embodiments, if there is only one candidate palmprint feature in the second palmprint feature library that meets the palmprint matching conditions, it means that there is a unique candidate palmprint feature that matches the palmprint feature to be detected. Therefore, the identity authentication result is set to pass palmprint recognition, and the identity identifier bound to this candidate palmprint feature is returned.

[0150] In the above process, if there is only one matching candidate palmprint feature in the second palmprint feature library, when the match fails in the first palmprint feature library, the only matching candidate palmprint feature can be found in the second palmprint feature library, thereby ensuring that the identity of the target object can be accurately queried, so as to maximize the success rate of the palmprint recognition service and avoid the damage to the experience caused by the target object repeatedly turning on palmprint recognition.

[0151] A3. If there is more than one candidate palm print feature, the palm scanning device determines the identity authentication result of the target object based on the candidate facial features bound to the candidate palm print features.

[0152] In some embodiments, if there is more than one candidate palmprint feature that meets the palmprint matching conditions in the second palmprint feature library, it means that there are multiple candidate palmprint features that match the palmprint feature to be tested, and the second palmprint feature library is already a full feature library. The most likely reason is that the quality of the captured palm image is not high. At this time, the identity authentication result can be directly set to fail palmprint recognition, prompting the target object to re-initiate palmprint recognition or switch the identity authentication method, thereby fully ensuring the security of the identity authentication service.

[0153] In other embodiments, the following steps A31 to A33 can also be used to introduce candidate facial features bound to each candidate palm print feature for auxiliary identification (or called facial verification), and the candidate palm print features can be secondary filtered based on the candidate facial features. In this way, the identity authentication result of the target object is determined based on the final filtering result, which can maximize the success rate of the palm print recognition service and avoid the damage to the experience caused by the target object repeatedly turning on palm print recognition.

[0154] A31. If there is more than one candidate palm print feature, the palm scanning device determines the palm print similarity difference between the candidate palm print feature with the greatest palm print similarity and each of the remaining candidate palm print features.

[0155] In some embodiments, if there is more than one candidate palm print feature that meets the palm print matching conditions in the second palm print feature library, the palm scanning device can determine the palm print similarity difference between the candidate palm print feature with the greatest palm print similarity and each of the remaining candidate palm print features, where each of the remaining candidate palm print features refers to the candidate palm print features that meet the palm print matching conditions, except for the candidate palm print feature with the greatest palm print similarity.

[0156] In one possible implementation, all candidate palmprint features matched in the second palmprint feature library are sorted in descending order of palmprint similarity, so that the candidate palmprint feature ranked first is the candidate palmprint feature with the greatest palmprint similarity. Next, the palmprint similarity difference between the candidate palmprint feature ranked first and each candidate palmprint feature ranked in the remaining order is calculated, so as to obtain at least one palmprint similarity difference, each palmprint similarity difference representing the difference in palmprint similarity between the candidate palmprint feature ranked first and another candidate palmprint feature in the ranking.

[0157] A32. If the similarity differences of the palm prints are not less than the difference threshold, the palm scanning device sets the identity authentication result as passed palm print recognition and returns the identity identifier bound to the candidate palm print feature with the greatest palm print similarity.

[0158] In some embodiments, it is determined whether the palm print similarity difference calculated in step A31 is less than a preset difference threshold, so that it can be evaluated whether the candidate palm print feature with the greatest palm print similarity and each of the other candidate palm print features are "highly similar features". If all the palm print similarity differences are not less than the difference threshold, it means that the candidate palm print feature with the greatest palm print similarity and each of the other candidate palm print features are not "highly similar features". Therefore, the risk of passing palm print recognition is relatively low at this time, and the identity authentication result is set to pass palm print recognition, and the identity identifier bound to the candidate palm print feature with the greatest palm print similarity is returned.

[0159] In the above process, by judging whether the palmprint similarity difference is less than the difference threshold, it is possible to evaluate whether the candidate palmprint feature with the greatest palmprint similarity and each of the other candidate palmprint features belong to "high similarity features". Therefore, if there are no "high similarity features", it can be relatively accurately considered that the candidate palmprint feature with the greatest palmprint similarity is the only match. The success rate of palmprint recognition can be guaranteed as much as possible under the premise of ensuring accurate identity authentication, avoiding the damage to the experience caused by prompting the target object to repeatedly initiate palmprint recognition.

[0160] A33. If the similarity difference of any palm print is less than the difference threshold, the palm scanning device determines the identity authentication result of the target object based on the candidate facial features bound to the candidate palm print features.

[0161] In some embodiments, it is determined whether the palm print similarity difference calculated in step A31 is less than a preset difference threshold, so that it can be evaluated whether the candidate palm print feature with the greatest palm print similarity and each of the other candidate palm print features are "high similarity features". If any palm print similarity difference is less than the difference threshold, it means that the candidate palm print feature with the greatest palm print similarity and at least one of the other candidate palm print features are "high similarity features". Therefore, it means that at least two matched candidate palm print features are highly similar. Then it may be impossible to accurately judge which identity identifier the target object belongs to. Palm print recognition will bring certain risks. In order to ensure the accuracy of the identity authentication service, the candidate facial features bound to each candidate palm print feature can be introduced through the following steps A33-1 to A33-4 for auxiliary identification (or facial verification). The candidate palm print features are secondary filtered based on the candidate facial features. In this way, the identity authentication result of the target object is determined based on the final filtering result, which can maximize the success rate of the palm print recognition service and avoid the damage to the experience caused by the target object repeatedly turning on palm print recognition.

[0162] A33-1. The palm scanning device determines the candidate palm print feature with the greatest palm print similarity and the candidate palm print feature with a palm print similarity difference less than the difference threshold as the target candidate palm print feature.

[0163] In some embodiments, since not all of the multiple candidate palmprint features screened out from the second palmprint feature library in step A3 have a palmprint similarity difference with the candidate palmprint feature with the greatest palmprint similarity that is less than the difference threshold, when considering the candidate facial features bound thereto, in order to further save palmprint recognition computational complexity and improve palmprint recognition efficiency, only the candidate palmprint feature with the greatest palmprint similarity itself and the candidate palmprint features screened out in step A33 whose palmprint similarity difference is less than the difference threshold need to be considered. For the sake of simplicity, the candidate palmprint feature with the greatest palmprint similarity and the candidate palmprint feature whose palmprint similarity difference is less than the difference threshold are determined as target candidate palmprint features. In this way, there is no need to query the candidate facial features bound to all candidate palmprint features, nor is there any need to involve all candidate facial features bound to candidate palmprint features in facial similarity judgment. Only the candidate facial features bound to the target candidate palmprint features need to be considered, and the candidate facial features bound to the target candidate palmprint features need to be involved in facial similarity judgment. This further compresses the computational complexity of facial assisted recognition, reduces computational overhead, shortens waiting delay, and improves identity authentication efficiency.

[0164] A33-2. The palm scanning device obtains the facial similarity between the candidate facial feature bound to each target candidate palm print feature and the pre-stored facial feature.

[0165] In some embodiments, the palm scanning device queries the association table described in step 303 for each target candidate palm print feature determined in step A33-1 to retrieve the candidate facial feature associated with the target candidate palm print feature. This process requires using the palm print feature ID of the target candidate palm print feature recorded in the Value record to retrieve the facial feature ID of the candidate facial feature associated with the target candidate palm print feature recorded in the Key record. This can be achieved using a hash table or other methods, and will not be further described.

[0166] In some embodiments, for each target candidate palm print feature, after the palm scanning device finds the facial feature ID of the candidate facial feature bound to the target candidate palm print feature, it takes out the candidate facial feature indicated by the facial feature ID from the facial feature library and calculates the facial similarity between the candidate facial feature and the pre-stored facial feature of the target object obtained in step 302. The method for calculating the facial similarity is described in step 302 and will not be repeated here.

[0167] A33-3. The palm scanning device eliminates target candidate palm print features whose facial similarity does not meet the facial matching condition, and obtains retained candidate palm print features. The facial matching condition indicates that the candidate facial feature and the pre-stored facial feature belong to the same object.

[0168] In some embodiments, after the palm-scanning device calculates the facial similarity between the facial feature to be tested and the candidate facial feature bound to each target candidate palm print feature one by one, it then determines whether the facial similarity is greater than a pre-set facial similarity threshold. If it is greater than the facial similarity threshold, it means that the candidate facial feature and the facial feature to be tested match, and both meet the facial matching conditions. If it is not greater than the facial similarity threshold, it means that the candidate facial feature and the facial feature to be tested do not match, and both do not meet the facial matching conditions, and the facial similarity between the next candidate facial feature and the facial feature to be tested is determined. The above operation is repeated until all candidate facial features bound to the target candidate palm print features are traversed. Then, all target candidate palm print features whose facial similarities do not meet the facial matching conditions are eliminated from the target candidate palm print features, and the target candidate palm print features remaining after elimination are called retained candidate palm print features.

[0169] In the above process, since the target candidate palm print features are already "highly similar features", the bound candidate facial features are introduced for auxiliary recognition. By calculating the facial similarity between the candidate facial features and the facial features to be tested, it is judged whether the facial matching conditions are met. In this way, if a candidate facial feature does not meet the facial matching conditions with the facial features to be tested, it means that there is a high probability that the two do not belong to the same object. The target candidate palm print features bound to the candidate facial features that do not meet the facial matching conditions are eliminated to obtain the retained candidate palm print features, which helps to further distinguish the "highly similar features" with very close palm prints and improve the accuracy and success rate of palm print recognition.

[0170] A33-4. The palm scanning device determines the identity authentication result of the target object based on the retained candidate palm print features.

[0171] In some embodiments, based on the retained candidate palmprint features obtained in step A33-4, the target object's identity authentication result is determined. Optionally, if there is only one retained candidate palmprint feature, the identity authentication result is set to pass palmprint recognition, and the identity identifier associated with the retained candidate palmprint feature is returned; if there are multiple retained candidate palmprint features, the identity authentication result is set to fail palmprint recognition, and the target object is prompted to re-initiate palmprint recognition or switch the identity authentication method.

[0172] In the above process, if there is only one retained candidate palmprint feature, it means that only the object to which the retained candidate palmprint feature belongs meets both the facial matching conditions and the palmprint matching conditions of the target object. Therefore, the identity of the target object can be uniquely identified. The identity authentication result is directly set as passed palmprint recognition, and the identity identifier bound to the retained candidate palmprint feature is returned, thereby ensuring the accuracy and success rate of palmprint recognition. Correspondingly, if there are multiple retained candidate palmprint features, it means that there are still multiple objects that match both the palm and face of the target object. This is an event with a very low probability, but if it occurs, it means that the quality of the captured palm image may be low. In order to ensure the security of the identity authentication service, it is directly returned that the palmprint recognition fails, prompting the target object to re-initiate palmprint recognition or switch the identity authentication method, which can fully ensure the security of the service.

[0173] In the above steps 306-308, a possible implementation method is provided for matching the palmprint feature to be detected with each palmprint feature stored in the first palmprint feature library, and determining the identity authentication result of the target object based on the matched palmprint features. That is, the first palmprint feature library is first used for priority comparison to ensure that there is a high probability of directly finding the matching palmprint feature in the first palmprint feature library, and the computing overhead of most palmprint recognition services is controlled within the range of the library search of the first palmprint feature library, thereby compressing the performance overhead of most identity recognition services. Only when the first palmprint feature library fails to match, the second palmprint feature library will be introduced for full library search to maximize the success rate and accuracy of palmprint recognition. This multi-level search method can effectively control the palmprint recognition delay within the range allowed by high real-time requirements, such as numerical value transfer services, greatly improving the palmprint recognition speed.

[0174] Figure 4 This is a schematic diagram of the principle of an identity authentication method provided in an embodiment of the present application, such as Figure 4 As shown, after the target object performs the identity authentication operation, it applies for camera permission. After the target object fully authorizes and agrees, the camera includes a color camera 410 and an infrared camera 420. The color camera 410 collects the color facial image and color palm image of the target object, and the infrared camera 420 collects the infrared facial image and infrared palm image of the target object. Then, the color facial image is used to extract color facial features, the color palm image is used to extract color palm print features, the infrared facial image is used to extract infrared facial features, and the infrared palm image is used to extract infrared palm print features. The color facial features and the infrared facial features can be fused to form the target object's facial features to be detected, and the color palm print features and the infrared palm print features can be fused to form the target object's palm print features to be detected. The facial features to be detected and the palm print features to be detected are used to perform a priority comparison in the first palm print feature library 430. If the comparison hits a matching palm print feature, the target object's identity is directly obtained. If the comparison does not hit a matching palm print feature (i.e., the first palm print feature library 430 fails to match), a full comparison is continued in the second palm print feature library 440 to obtain the target object's identity.

[0175] In other embodiments, when the identity authentication result indicates that palm print recognition has been passed, the target object can also be verified for liveness using the infrared palm image and infrared facial image of the target object. Only when both palm print recognition and liveness verification have been passed is the overall identity authentication considered to have passed, and the identity identifier of the target object is returned, thereby preventing attackers from using certain photos to impersonate the target object, further improving the security of identity authentication.

[0176] Figure 5This is a schematic diagram of another identity authentication method provided by an embodiment of the present application. Figure 5 As shown, after the target object performs the identity authentication operation, it applies for camera permission. After the target object fully authorizes and agrees, the camera module 510 (i.e., the camera, which may include a color camera + an infrared camera) collects the target object's color facial image, color palm image, infrared facial image, and infrared palm image. Next, the algorithm image optimization module 520 performs image optimization on the above-mentioned color facial image, color palm image, infrared facial image, and infrared palm image, and uses the image optimization algorithm to filter out seriously unqualified images to prevent unqualified images from being sent for palmprint recognition, resulting in a waste of computing resources, while also avoiding a reduction in the palmprint recognition error rate. Next, a first palmprint feature library 531 is maintained. The first palmprint feature library 531 uses the correlation between facial features and palmprint features to store the palmprint features of each object that has recently used the identity authentication service. After extracting the palmprint features to be checked of the target object, the palmprint features to be checked are preferentially sent to the first palmprint feature library 531 for priority comparison. If a matching palmprint feature is found (and there is no highly similar feature), the infrared facial image and the infrared palm image are sent to the liveness verification module 540 for liveness verification. If the liveness verification is passed, the identity database 550 is queried to find the identity bound to the matching palmprint feature, the identity authentication result is configured as passing palmprint recognition, and the identity is returned. If the liveness verification is not passed, the identity authentication result is configured as failing palmprint recognition, and the target object is prompted to re-initiate palmprint recognition or switch the identity authentication method. In addition, if the palm print feature to be checked fails to match in the first palm print feature library 531, the palm print feature to be checked will be sent to the second palm print feature library 532 for full library comparison. If a matching palm print feature is found (and there is no highly similar feature), the infrared facial image and the infrared palm image will be sent to the liveness verification module 540 for liveness verification. If the liveness verification is passed, the identity identification database 550 will be queried to find the identity identification bound to the matching palm print feature, and the identity authentication result will be configured to pass palm print recognition and return the identity identification; if the liveness verification is not passed, the identity authentication result will be configured as failed palm print recognition, and the target object will be prompted to re-initiate palm print recognition or switch the identity authentication method. Furthermore, in order to improve the palm print recognition speed of the palm-scanning devices deployed in various places, the first palm print feature library 531 and / or the second palm print feature library 532 can also be synchronized between the palm-scanning devices. When synchronizing across devices, feature synchronization can be achieved based on D2D communication or by using a cloud server. No specific limitation is made here. This can ensure that the first palm print feature library 531 and / or the second palm print feature library 532 is updated in time between different palm-scanning devices, thereby improving the palm print recognition speed of all palm-scanning devices.

[0177] All the above optional technical solutions can be arbitrarily combined to form optional embodiments of the present disclosure, and will not be described in detail here.

[0178] The method provided in the embodiment of the present application utilizes the features stored in the first palmprint feature library for priority palmprint matching, so that the full set of palmprint features does not need to be stored in the first palmprint feature library. After the pre-stored palmprint features are loaded into the first palmprint feature library through the association between the pre-stored facial features and the pre-stored palmprint features, since the palmprint recognition process does not require matching all the features in the library one by one, it is equivalent to using the facial image to narrow the palmprint matching range, thereby increasing the probability of successfully matching the palmprint features to be tested using the first palmprint feature library, greatly reducing the computational overhead of the palmprint recognition process, and improving the efficiency of identity authentication based on palmprint recognition. Furthermore, since the pre-stored facial features are used to lock the pre-stored palmprint features, the matching palmprint features found in the first palmprint feature library are more likely to be the pre-stored palmprint features, thereby also improving the success rate of identity authentication based on palmprint recognition.

[0179] Next, we will combine Figure 6 , the binding process of the facial features and palm print features of the target object in the embodiment of the present application is described. Figure 6 This is a flowchart of binding facial features and palm print features provided by an embodiment of the present application, such as Figure 6 As shown, after the target object authorizes the registration of the identity authentication service and grants the camera permission, when the target object approaches the palm swiping device, the camera will take advantage of the large FoV (Field of View) range to capture the target object's color facial image and infrared facial image before the target object performs the palm extension operation. The target object's color facial features and infrared facial features are then extracted. The color facial features and infrared facial features are fused to obtain the target object's facial features to be tested. The facial features to be tested are used as the pre-stored facial features entered when the target object was registered, and the pre-stored facial features are added to the facial feature library. Then, after the target object begins to swipe the palm, i.e., perform the palm extension operation, the target object's color palm image and infrared palm image are captured. The target object's color palm print features and infrared palm print features are then extracted. The color palm print features and infrared palm print features are fused to obtain the target object's palm print features to be tested. A matching pre-stored palm print feature is found in the second palm print feature library. Then, an association relationship is established between the pre-stored facial features and the pre-stored palm print features, that is, the pre-stored facial features and the pre-stored palm print features of the target object are bound.

[0180] Next, we will combine Figure 7 , the process of using the first palmprint feature library for priority comparison in the embodiment of the present application is described. Figure 7 This is a flowchart of a method of using the first palmprint feature library for priority comparison provided by the embodiment of the present application, such as Figure 7 As shown, after the target person authorizes the identity authentication service and grants camera access, when the target person approaches the palm scanning device, the camera, taking advantage of its large Field of View (FoV), captures both the target person's color and infrared facial images before the target person extends their palm. The camera then extracts the target person's color and infrared facial features, merging them to obtain the target person's facial features to be tested. The system then searches the facial feature library for matching pre-stored facial features based on the facial features to be tested. The associated relationship table is used to retrieve the pre-stored palmprint features bound to the pre-stored facial features, and these pre-stored palmprint features are added to the first palmprint feature library. Next, after the target object starts to swipe its palm, that is, performs the palm extension operation, the color palm image and infrared palm image of the target object are collected, and then the color palm print features and infrared palm print features of the target object are extracted. The color palm print features and infrared palm print features are fused to obtain the palm print features to be detected of the target object. After that, palm print matching is performed preferentially from the first palm print feature library. If a matching palm print feature is found in the first palm print feature library, the identity of the target object can be directly returned without being sent to the second palm print feature library. If no matching palm print feature is found in the first palm print feature library, full palm print matching is continued in the second palm print feature library. If a matching palm print feature is found in the second palm print feature library, the identity of the target object is returned. If no matching palm print feature is found in the second palm print feature library, the target object is prompted to register for the identity authentication service, and the pre-stored palm print features and pre-stored facial features are entered in sequence, and the pre-stored facial features and pre-stored palm print features of the target object are bound. In this way, the palmprint recognition speed of the target object can be improved without affecting the security of palmprint recognition through the hierarchical query architecture constructed by the first palmprint feature library and the second palmprint feature library.

[0181] Next, we will combine Figure 8 , the process of using facial features for auxiliary recognition in the embodiment of the present application is described. Figure 8 This is a flowchart of an embodiment of the present application that uses facial features for auxiliary recognition, such as Figure 8 As shown, in Figure 7As shown, based on the priority comparison using the first palmprint feature library, it can be seen that if multiple matching candidate palmprint features are found in the second palmprint feature library, and there are "high similarity features" between these multiple candidate palmprint features, in traditional cases, recognition will generally be directly rejected because high similarity features are more prone to misidentification, affecting the user experience. Usually, an error prompt will be directly returned, requiring the target object to re-initiate palmprint recognition. However, in the embodiment of the present application, facial features are used for auxiliary identification. If multiple matching candidate palmprint features are found in the second palmprint feature library, and there are "highly similar features" between these multiple candidate palmprint features, the candidate facial features bound to each candidate palmprint feature are used to match them with the pre-stored facial features of the target object, and it is determined whether the facial features bound to these matched palmprint features are also similar to the pre-stored facial features. If the facial features are not similar, the palmprint features bound to these dissimilar facial features are directly filtered out. In this way, the "highly similar features" can be distinguished in most cases, and the identity identifier of the target object to which the palmprint features finally matched is found, thereby improving the palmprint recognition speed of the target object and ensuring the success rate of palmprint recognition without affecting the security of palmprint recognition.

[0182] Figure 9 This is a structural diagram of an identity authentication device provided in an embodiment of the present application. Figure 9 As shown, the device includes:

[0183] A feature determination module 901 is configured to determine, in response to an identity authentication operation initiated by a target object, pre-stored facial features of the target object based on a captured facial image of the target object;

[0184] An adding module 902 is configured to add the pre-stored palmprint feature bound to the pre-stored facial feature to a first palmprint feature library, the first palmprint feature library being used to store features that are prioritized for palmprint matching;

[0185] An extraction module 903 is configured to extract the palmprint features of the target object based on the acquired palm image of the target object;

[0186] The result determination module 904 is configured to match the palmprint feature to be detected with each palmprint feature stored in the first palmprint feature library, and determine the identity authentication result of the target object based on the matched palmprint features.

[0187] The device provided by the embodiment of the present application utilizes the features stored in the first palmprint feature library for priority palmprint matching, so that the full set of palmprint features does not need to be stored in the first palmprint feature library. After the pre-stored palmprint features are loaded into the first palmprint feature library through the association between the pre-stored facial features and the pre-stored palmprint features, since the palmprint recognition process does not need to match the full set of features one by one, it is equivalent to using the facial image to narrow the palmprint matching range, thereby improving the probability of successfully matching the palmprint features to be tested using the first palmprint feature library, greatly reducing the computational overhead of the palmprint recognition process, and improving the efficiency of identity authentication based on palmprint recognition. Furthermore, since the pre-stored facial features are used to lock the pre-stored palmprint features, the matching palmprint features found in the first palmprint feature library are more likely to be the pre-stored palmprint features, thereby also improving the success rate of identity authentication based on palmprint recognition.

[0188] In some embodiments, based on Figure 9 The result determination module 904 includes:

[0189] an acquisition submodule, configured to acquire a palmprint similarity between the palmprint feature to be detected and each palmprint feature stored in the first palmprint feature library;

[0190] a result determination submodule for determining an identity authentication result of the target object based on the matched palmprint feature if a palmprint feature whose palmprint similarity meets a palmprint matching condition exists in the first palmprint feature library, wherein the palmprint matching condition indicates that the palmprint feature to be detected and the palmprint feature belong to the same object;

[0191] The query and determination submodule is used to query palmprint features whose palmprint similarity meets the palmprint matching conditions from the second palmprint feature library if there is no palmprint feature whose palmprint similarity meets the palmprint matching conditions in the first palmprint feature library, and determine the identity authentication result of the target object based on the queried palmprint features. The second palmprint feature library is used to store the full set of features for palmprint matching.

[0192] In some embodiments, based on Figure 9 The query determination submodule includes:

[0193] a query unit, configured to query, based on the palmprint feature to be detected, candidate palmprint features whose palmprint similarity meets the palmprint matching condition from the second palmprint feature database;

[0194] A result returning unit is configured to, if there is only one candidate palmprint feature, set the identity authentication result as passed palmprint recognition and return the identity identifier bound to the candidate palmprint feature;

[0195] The result determination unit is configured to determine the identity authentication result of the target object based on the candidate facial features bound to the candidate palmprint features if there is more than one candidate palmprint feature.

[0196] In some embodiments, based on Figure 9 The result determination unit comprises:

[0197] a determination subunit, configured to determine, if there is more than one candidate palmprint feature, a palmprint similarity difference between the candidate palmprint feature with the greatest palmprint similarity and each of the remaining candidate palmprint features;

[0198] The result return subunit is used to set the identity authentication result as passed palmprint recognition if the similarity difference of each palmprint is not less than the difference threshold, and return the identity identifier bound to the candidate palmprint feature with the greatest palmprint similarity;

[0199] The result determination subunit is configured to determine the identity authentication result of the target object based on the candidate facial features bound to the candidate palmprint features if the similarity difference of any palmprint is less than the difference threshold.

[0200] In some embodiments, based on Figure 9 The device is composed of the result determining subunit including:

[0201] The first determining sub-subunit is configured to determine the candidate palmprint feature with the greatest palmprint similarity and the candidate palmprint feature with a palmprint similarity difference less than a difference threshold as the target candidate palmprint feature;

[0202] An acquisition sub-subunit, configured to acquire a facial similarity between the candidate facial feature bound to each target candidate palmprint feature and the pre-stored facial feature;

[0203] a rejection sub-subunit for rejecting target candidate palmprint features whose facial similarity does not meet a facial matching condition, thereby obtaining retained candidate palmprint features, wherein the facial matching condition indicates that the candidate facial feature and the pre-stored facial feature belong to the same object;

[0204] The second determining sub-subunit is configured to determine an identity authentication result of the target object based on the retained candidate palmprint features.

[0205] In some embodiments, the second determining sub-subunit is configured to:

[0206] If there is only one candidate palmprint feature left, set the identity authentication result as passed palmprint recognition and return the identity identifier bound to the candidate palmprint feature;

[0207] If there are multiple candidate palmprint features left, the identity authentication result is set to fail palmprint recognition.

[0208] In some embodiments, the feature determination module 901 is used to:

[0209] Extracting facial features to be detected of the target object based on the collected facial image of the target object;

[0210] The facial features to be detected are matched with various facial features stored in a facial feature library, and the matched facial features are determined as pre-stored facial features of the target object.

[0211] In some embodiments, the result determination module 904 is further configured to:

[0212] If no matching facial features are found in the facial feature library, the palmprint features to be checked are matched with the palmprint features stored in the second palmprint feature library, and the identity authentication result of the target object is determined based on the matched palmprint features. The second palmprint feature library is used to store all the features of the palmprint match;

[0213] The facial feature to be detected is used as the pre-stored facial feature of the target object, and the pre-stored facial feature is bound to the palm print feature matched in the second palm print feature library.

[0214] In some embodiments, based on Figure 9 The device is composed of:

[0215] The prompting module is used to prompt the target object to retake the facial image or palm image if the collected facial image or palm image of the target object does not meet the image detection conditions.

[0216] In some embodiments, based on Figure 9 The device is composed of:

[0217] The synchronization module is used to synchronize the palmprint features stored in the first palmprint feature library in response to the data synchronization instruction of the first palmprint feature library.

[0218] In some embodiments, based on Figure 9 The device is composed of:

[0219] The acquisition module is used to acquire a facial image of the target object in response to the target object performing an identity authentication operation; and to acquire a palm image of the target object in response to the target object performing a palm extension operation.

[0220] All the above optional technical solutions can be arbitrarily combined to form optional embodiments of the present disclosure, and will not be described in detail here.

[0221] It should be noted that the identity authentication device provided in the above embodiment only illustrates the division of the aforementioned functional modules when authenticating the identity of a target object. In actual applications, the aforementioned functions can be assigned to different functional modules as needed, that is, the internal structure of the computer device can be divided into different functional modules to complete all or part of the functions described above. In addition, the identity authentication device provided in the above embodiment and the identity authentication method embodiment are based on the same concept. The specific implementation process is detailed in the identity authentication method embodiment and will not be repeated here.

[0222] Figure 10 1 is a schematic diagram of the structure of a computer device provided in an embodiment of the present application. The computer device can be implemented as the palm-swiping device provided in the above embodiments, or other terminals supporting palmprint recognition functions. Here, the palm-swiping device 1000 is used as an example for illustration. Optionally, the palm-swiping device 1000 may include, but is not limited to, a palm-swiping payment terminal, a smartphone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smartwatch, and the like. The palm-swiping device 1000 may also be referred to as a user device, a portable terminal, a laptop terminal, a desktop terminal, and other names.

[0223] Typically, the palm-brushing device 1000 includes a processor 1001 and a memory 1002 .

[0224] Optionally, the processor 1001 includes one or more processing cores, such as a 4-core processor, an 8-core processor, etc. Optionally, the processor 1001 is implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), or PLA (Programmable Logic Array). In some embodiments, the processor 1001 includes a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in the standby state. In some embodiments, the processor 1001 is integrated with a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 1001 also includes an AI (Artificial Intelligence) processor, which is used to process computing operations related to machine learning.

[0225] In some embodiments, the memory 1002 includes one or more computer-readable storage media, optionally, the computer-readable storage medium is non-transitory. Optionally, the memory 1002 also includes a high-speed random access memory and a non-volatile memory, such as one or more disk storage devices, flash memory storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory 1002 is used to store at least one program code, which is used to be executed by the processor 1001 to implement the identity authentication method provided in each embodiment of the present application.

[0226] In some embodiments, the palm-brush device 1000 may optionally include a peripheral device interface 1003 and at least one peripheral device. The processor 1001, memory 1002, and peripheral device interface 1003 may be connected via a bus or signal lines. Each peripheral device may be connected to the peripheral device interface 1003 via a bus, signal lines, or circuit boards. Specifically, the peripheral device may include at least one of a radio frequency circuit 1004, a display screen 1005, a camera assembly 1006, an audio circuit 1007, and a power supply 1008.

[0227] The peripheral device interface 1003 can be used to connect at least one I / O (Input / Output)-related peripheral device to the processor 1001 and the memory 1002. In some embodiments, the processor 1001, the memory 1002, and the peripheral device interface 1003 are integrated on the same chip or circuit board; in some other embodiments, any one or two of the processor 1001, the memory 1002, and the peripheral device interface 1003 are implemented on separate chips or circuit boards, which is not limited in this embodiment.

[0228] The RF circuit 1004 is used to receive and transmit RF (Radio Frequency) signals, also known as electromagnetic signals. The RF circuit 1004 communicates with communication networks and other communication devices via electromagnetic signals. The RF circuit 1004 converts electrical signals into electromagnetic signals for transmission, or converts received electromagnetic signals into electrical signals. Optionally, the RF circuit 1004 includes: an antenna system, an RF transceiver, one or more amplifiers, a tuner, an oscillator, a digital signal processor, a codec chipset, a user identity module card, and the like. Optionally, the RF circuit 1004 communicates with other palm-swipe devices via at least one wireless communication protocol. Such wireless communication protocols include, but are not limited to, metropolitan area networks, various generations of mobile communication networks (2G, 3G, 4G, and 5G), wireless local area networks, and / or WiFi (Wireless Fidelity) networks. In some embodiments, the RF circuit 1004 also includes circuitry related to NFC (Near Field Communication), which is not limited in this application.

[0229] The display screen 1005 is used to display a UI (User Interface). Optionally, the UI includes graphics, text, icons, videos, and any combination thereof. When the display screen 1005 is a touch screen display, the display screen 1005 also has the ability to collect touch signals on the surface or above the surface of the display screen 1005. The touch signal can be input as a control signal to the processor 1001 for processing. Optionally, the display screen 1005 is also used to provide virtual buttons and / or virtual keyboards, also known as soft buttons and / or soft keyboards. In some embodiments, there is one display screen 1005, which is provided on the front panel of the palm-brushing device 1000; in other embodiments, there are at least two display screens 1005, which are respectively provided on different surfaces of the palm-brushing device 1000 or are foldable in design; in some embodiments, the display screen 1005 is a flexible display screen, which is provided on a curved surface or a foldable surface of the palm-brushing device 1000. Even, optionally, the display screen 1005 is provided in a non-rectangular irregular shape, that is, a special-shaped screen. Optionally, the display screen 1005 is made of materials such as LCD (Liquid Crystal Display) and OLED (Organic Light-Emitting Diode).

[0230] The camera assembly 1006 is used to capture images or videos. Optionally, the camera assembly 1006 includes a front camera and a rear camera. Typically, the front camera is provided on the front panel of the palm-brushing device, and the rear camera is provided on the back of the palm-brushing device. In some embodiments, there are at least two rear cameras, which are any one of a main camera, a depth-of-field camera, a wide-angle camera, and a telephoto camera, so as to realize the fusion of the main camera and the depth-of-field camera to realize the background blur function, the fusion of the main camera and the wide-angle camera to realize panoramic shooting and VR (Virtual Reality) shooting function or other fusion shooting functions. In some embodiments, the camera assembly 1006 also includes a flash. Optionally, the flash is a monochrome temperature flash, or a dual-color temperature flash. A dual-color temperature flash refers to a combination of a warm light flash and a cold light flash, which is used for light compensation under different color temperatures.

[0231] In some embodiments, the audio circuit 1007 includes a microphone and a speaker. The microphone is used to collect sound waves from the user and the environment, and convert the sound waves into electrical signals and input them into the processor 1001 for processing, or input them into the radio frequency circuit 1004 to achieve voice communication. For the purpose of stereo acquisition or noise reduction, there are multiple microphones, which are respectively arranged at different parts of the palm-brushing device 1000. Optionally, the microphone is an array microphone or an omnidirectional acquisition microphone. The speaker is used to convert the electrical signal from the processor 1001 or the radio frequency circuit 1004 into sound waves. Optionally, the speaker is a traditional thin film speaker or a piezoelectric ceramic speaker. When the speaker is a piezoelectric ceramic speaker, it can not only convert the electrical signal into sound waves audible to humans, but also convert the electrical signal into sound waves inaudible to humans for purposes such as ranging. In some embodiments, the audio circuit 1007 also includes a headphone jack.

[0232] Power supply 1008 is used to power various components of palm-brush device 1000. Optionally, power supply 1008 is AC power, DC power, a disposable battery, or a rechargeable battery. When power supply 1008 includes a rechargeable battery, the rechargeable battery supports wired charging or wireless charging. The rechargeable battery is also configured to support fast charging technology.

[0233] In some embodiments, the palm-brushing device 1000 further includes one or more sensors 1010 . The one or more sensors 1010 include, but are not limited to, an optical sensor 1011 and a proximity sensor 1012 .

[0234] Optical sensor 1011 is used to detect ambient light intensity. In one embodiment, processor 1001 controls the display brightness of display screen 1005 based on the ambient light intensity detected by optical sensor 1011. Specifically, when the ambient light intensity is high, the display brightness of display screen 1005 is increased; when the ambient light intensity is low, the display brightness of display screen 1005 is decreased. In another embodiment, processor 1001 also dynamically adjusts the capture parameters of camera assembly 1006 based on the ambient light intensity detected by optical sensor 1011.

[0235] The proximity sensor 1012, also known as a distance sensor, is typically located on the front panel of the palm-brushing device 1000. The proximity sensor 1012 is used to detect the distance between the user and the front of the palm-brushing device 1000. In one embodiment, when the proximity sensor 1012 detects that the distance between the user and the front of the palm-brushing device 1000 is gradually decreasing, the processor 1001 controls the display screen 1005 to switch from the screen-on state to the screen-off state. When the proximity sensor 1012 detects that the distance between the user and the front of the palm-brushing device 1000 is gradually increasing, the processor 1001 controls the display screen 1005 to switch from the screen-off state to the screen-on state.

[0236] Those skilled in the art will understand that Figure 10 The structure shown in the figure does not constitute a limitation on the palm-brushing device 1000, and it can include more or fewer components than shown in the figure, or combine certain components, or adopt a different component arrangement.

[0237] In an exemplary embodiment, a computer-readable storage medium is also provided, such as a memory including at least one computer program. The at least one computer program can be executed by a processor in a computer device to perform the identity authentication method described in each of the above embodiments. For example, the computer-readable storage medium includes ROM (Read-Only Memory), RAM (Random-Access Memory), CD-ROM (Compact Disc Read-Only Memory), magnetic tape, floppy disk, and optical data storage device.

[0238] In an exemplary embodiment, a computer program product is also provided, comprising one or more computer programs stored in a computer-readable storage medium. One or more processors of a computer device can read the one or more computer programs from the computer-readable storage medium and execute the one or more computer programs, thereby enabling the computer device to perform the identity authentication method of the above-described embodiment.

[0239] Those skilled in the art will understand that all or part of the steps to implement the above embodiments can be completed by hardware, or can be completed by instructing relevant hardware through a program. Optionally, the program is stored in a computer-readable storage medium. Optionally, the above-mentioned storage medium is a read-only memory, a disk or an optical disk, etc.

[0240] The above description is merely an optional embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. An identity authentication method, characterized in that: The method comprises: In response to an identity authentication operation initiated by a target object, determining pre-stored facial features of the target object based on a captured facial image of the target object; Based on the pre-stored facial features, determining a pre-stored palm print feature bound to the pre-stored facial features, wherein a connection relationship has been pre-established between the pre-stored palm print feature and the pre-stored facial features; adding the pre-stored palmprint features to a first palmprint feature library, wherein the first palmprint feature library is used to store features that are prioritized for palmprint matching; In response to the target object performing a palm extension operation, collecting a palm image of the target object, and extracting a palmprint feature to be detected of the target object based on the collected palm image of the target object; Obtaining palmprint similarity between the palmprint feature to be detected and each palmprint feature stored in the first palmprint feature library; If a palmprint feature exists in the first palmprint feature database whose palmprint similarity meets a palmprint matching condition, determining an identity authentication result of the target object based on the matched palmprint feature, wherein the palmprint matching condition indicates that the palmprint feature to be detected and the palmprint feature belong to the same object; If there is no palmprint feature whose palmprint similarity meets the palmprint matching condition in the first palmprint feature library, querying a second palmprint feature library for candidate palmprint features whose palmprint similarity meets the palmprint matching condition based on the palmprint feature to be detected, wherein the second palmprint feature library is used to store all features for palmprint matching; If there is only one candidate palmprint feature, setting the identity authentication result to pass palmprint recognition and returning the identity identifier bound to the candidate palmprint feature; In the case where there is more than one candidate palmprint feature, determining a palmprint similarity difference between the candidate palmprint feature with the greatest palmprint similarity and each of the remaining candidate palmprint features, wherein each of the remaining candidate palmprint features refers to a candidate palmprint feature other than the candidate palmprint feature with the greatest palmprint similarity among the candidate palmprint features meeting the palmprint matching condition; If the similarity differences of all palmprints are not less than the difference threshold, the identity authentication result is set as passed palmprint recognition, and the identity identifier bound to the candidate palmprint feature with the greatest palmprint similarity is returned; if the similarity difference of any palmprint is less than the difference threshold, the identity authentication result of the target object is determined based on the candidate facial features bound to the candidate palmprint features; In the case that the candidate palmprint feature is not found, the identity authentication result is set as palmprint recognition failure, and the target object is prompted to re-initiate palmprint recognition or switch the identity authentication method.

2. The method according to claim 1, characterized in that The determining of the identity authentication result of the target object based on the candidate facial features bound to the candidate palmprint features includes: Determine the candidate palmprint feature with the greatest palmprint similarity and the candidate palmprint feature with a palmprint similarity difference less than the difference threshold as the target candidate palmprint feature; Obtaining facial similarity between the candidate facial features bound to each target candidate palmprint feature and the pre-stored facial features; Eliminating target candidate palmprint features whose facial similarity does not meet a facial matching condition to obtain retained candidate palmprint features, wherein the facial matching condition indicates that the candidate facial features and the pre-stored facial features belong to the same object; An identity authentication result of the target object is determined based on the retained candidate palmprint features.

3. The method according to claim 2, characterized in that Determining the identity authentication result of the target object based on the retained candidate palmprint features includes: If there is only one candidate palmprint feature left, the identity authentication result is set to pass palmprint recognition, and the identity identifier bound to the candidate palmprint feature is returned; If there are multiple candidate palmprint features left, the identity authentication result is set to fail palmprint recognition.

4. The method according to claim 1, wherein The determining of pre-stored facial features of the target object based on the collected facial image of the target object includes: Extracting facial features to be detected of the target object based on the collected facial image of the target object; The facial features to be detected are matched with various facial features stored in a facial feature library, and the matched facial features are determined as pre-stored facial features of the target object.

5. The method according to claim 4, characterized in that After extracting the palmprint features to be detected of the target object, the method further includes: If no matching facial features are found in the facial feature library, matching the palmprint features to be detected with each palmprint feature stored in a second palmprint feature library, and determining an identity authentication result of the target object based on the matched palmprint features, wherein the second palmprint feature library is used to store all palmprint matching features; The facial features to be detected are used as pre-stored facial features of the target object, and the pre-stored facial features are bound to the palm print features matched in the second palm print feature library.

6. The method according to claim 1, characterized in that The method further comprises: If the collected facial image or palm image of the target object does not meet the image detection conditions, the target object is prompted to retake the facial image or palm image.

7. The method according to claim 1, characterized in that The method further comprises: In response to the data synchronization instruction of the first palmprint feature library, data synchronization is performed on the palmprint features stored in the first palmprint feature library.

8. The method according to claim 1, characterized in that The method further comprises: In response to the target object performing an identity authentication operation, a facial image of the target object is collected.

9. An identity authentication device, characterized in that: The device comprises: a feature determination module, configured to determine, in response to an identity authentication operation initiated by a target object, pre-stored facial features of the target object based on a captured facial image of the target object; an adding module configured to determine, based on the pre-stored facial features, a pre-stored palmprint feature bound to the pre-stored facial features, wherein a connection relationship has been pre-established between the pre-stored palmprint feature and the pre-stored facial features; and add the pre-stored palmprint feature bound to the pre-stored facial features to a first palmprint feature library, wherein the first palmprint feature library is configured to store features that are prioritized for palmprint matching; an extraction module, configured to collect a palm image of the target object in response to the target object performing a palm extension operation, and extract a palmprint feature to be detected of the target object based on the collected palm image of the target object; The result determination module is used to obtain the palmprint similarity between the palmprint feature to be detected and each palmprint feature stored in the first palmprint feature library; if there is a palmprint feature in the first palmprint feature library whose palmprint similarity meets the palmprint matching condition, the identity authentication result of the target object is determined based on the matched palmprint feature, and the palmprint matching condition indicates that the palmprint feature to be detected and the palmprint feature belong to the same object; if there is no palmprint feature in the first palmprint feature library whose palmprint similarity meets the palmprint matching condition, based on the palmprint feature to be detected, a candidate palmprint feature whose palmprint similarity meets the palmprint matching condition is queried from the second palmprint feature library, and the second palmprint feature library is used to store all the features of palmprint matching; when there is only one candidate palmprint feature, the identity authentication result is set to pass palmprint recognition, and the identity bound to the candidate palmprint feature is returned. identification; when there is more than one candidate palmprint feature, determining the palmprint similarity difference between the candidate palmprint feature with the greatest palmprint similarity and each of the remaining candidate palmprint features, wherein each of the remaining candidate palmprint features refers to the candidate palmprint features other than the candidate palmprint feature with the greatest palmprint similarity among the candidate palmprint features that meet the palmprint matching conditions; if the palmprint similarity differences of all palmprints are not less than the difference threshold, setting the identity authentication result as passed palmprint recognition, and returning the identity identifier bound to the candidate palmprint feature with the greatest palmprint similarity; if any palmprint similarity difference is less than the difference threshold, determining the identity authentication result of the target object based on the candidate facial features bound to the candidate palmprint features; if the candidate palmprint feature is not found, setting the identity authentication result as failed palmprint recognition, and prompting the target object to re-initiate palmprint recognition or switch the identity authentication method.

10. A computer device, characterized in that: The computer device includes one or more processors and one or more memories, wherein at least one computer program is stored in the one or more memories, and the at least one computer program is loaded and executed by the one or more processors to implement the identity authentication method according to any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that The computer-readable storage medium stores at least one computer program, and the at least one computer program is loaded and executed by the processor to implement the identity authentication method according to any one of claims 1 to 8.

12. A computer program product, characterized in that The computer program product includes at least one computer program, and the at least one computer program is loaded and executed by a processor to implement the identity authentication method according to any one of claims 1 to 8.

Images