Data processing method and device, electronic equipment, storage medium and program product

Abstract

The embodiment of the invention discloses a data processing method and device, electronic equipment, a storage medium and a program product. According to the embodiment of the invention, a target single feature and a target account can be obtained; the association condition of the target multiple features and the target account in a database is determined, the target multiple features are multiple features corresponding to the target single features, and the database comprises the association relationship between the account and each single feature in the multiple features; if the target multi-feature and the target account are in a first association condition in the database, associating the target single feature with the target account; and if the target multi-feature and the target account are in a second association condition in the database, refusing to associate the target single feature and the target account. Therefore, according to the scheme, a multi-feature account association scene is specially designed, the problem that the same user uses multiple features to associate multiple accounts can be solved, one person has one account, and the authenticity and safety of the accounts are guaranteed.

Description

TECHNICAL FIELD

[0001] The present application relates to the computer field, and in particular to a data processing method and device, electronic equipment, storage medium and program product. BACKGROUND

[0002] Biometric identification technology is a technology that uses individual biological characteristics for identity authentication and recognition. It collects and analyzes individual physiological or behavioral characteristics, such as fingerprints, palm prints, faces, irises, voice prints, and other biological characteristics, and compares them with biological characteristics previously stored in the system to identify the individual's identity.

[0003] Biological characteristics can be divided into unique characteristics and multiple characteristics. Unique characteristics refer to biological identification characteristics in a unique body part of an object, such as facial features. Multiple characteristics refer to biological identification characteristics of at least two independent parts of the same body part of an object, such as left and right eye irises, left and right hand palm prints, and ten finger fingerprints.

[0004] By associating a user's unique characteristics with the user's account, the problem of a user registering or binding multiple accounts can be avoided, and one person, one account ensures the authenticity and security of the account. However, this solution is not suitable for multiple characteristics. If multiple characteristics are applied to account association, the problem of a user associating multiple accounts will occur, for example, one account of the user is associated with the user's left hand palm print, and another account of the user is associated with the user's right hand palm print, resulting in the user being associated with two accounts. SUMMARY

[0005] The embodiments of the present application provide a data processing method and device, electronic equipment, storage medium and program product, which can solve the problem of a user associating multiple accounts using multiple characteristics, and achieve one person, one account, ensuring the authenticity and security of the account.

[0006] The embodiments of the present application provide a data processing method, comprising:

[0007] Obtaining a target single characteristic and a target account, the target single characteristic being a single characteristic to be associated with the target account, the single characteristic being one of a multiple characteristic, the multiple characteristic being biological identification characteristics of at least two independent parts of the same body part of an object;

[0008] Determining the association of the target multiple characteristic and the target account in a database, the target multiple characteristic being a multiple characteristic corresponding to the target single characteristic, the database including the association relationship between the account and each single characteristic in the multiple characteristic;

[0009] If the target multiple characteristic and the target account are in a first association in the database, then associate the target single characteristic with the target account;

[0010] If the target multiple features and the target account are in the second association condition in the database, the association of the target single feature and the target account is rejected.

[0011] Embodiments of the present application also provide a data processing apparatus, comprising:

[0012] The acquisition unit is configured to acquire target single features and a target account, the target single features being single features to be associated with the target account, the single features being one of multiple features, the multiple features being biological recognition features of at least two independent parts of a same body part of an object;

[0013] The determination unit is configured to determine an association condition of the target multiple features and the target account in the database, the target multiple features being multiple features corresponding to the target single features, the database comprising an association relationship between an account and each single feature in the multiple features;

[0014] The association unit is configured to associate the target single features and the target account if the target multiple features and the target account are in the first association condition in the database.

[0015] The rejection unit is configured to reject the association of the target single features and the target account if the target multiple features and the target account are in the second association condition in the database.

[0016] In some embodiments, the determination unit comprises:

[0017] The detection subunit is configured to detect the association relationship of the target multiple features and the association relationship of the target account in the database respectively.

[0018] The first subunit is configured to determine that the target multiple features and the target account are in the first association condition in the database if there is an association relationship between the target multiple features and the target account in the database, or if there is no target multiple features and the target account in the database.

[0019] The second subunit is configured to determine that the target multiple features and the target account are in the second association condition in the database if there is an association relationship between the target multiple features and other accounts in the database, or if there is an association relationship between other multiple features and the target account, the other accounts being other accounts except the target account, and the other multiple features being other multiple features except the target multiple features.

[0020] In some embodiments, the determination unit is further configured to:

[0021] perform a fast search on the target single features in the database, so as to obtain a fast search result of the target single features in the database.

[0022] If the quick search results indicate that the target's single feature does not exist in the database, then the target's multiple features and the target account are determined to be in the first association situation in the database.

[0023] In some embodiments, the detection subunit is configured to:

[0024] Determine the similarity between a single feature of the target and each single feature in the database;

[0025] Based on similarity, identify similar features in the database that are similar to the target's single feature;

[0026] Identify accounts associated with similar characteristics;

[0027] Identify a single characteristic in the database that is associated with the target account.

[0028] In some embodiments, determining the similarity between a target single feature and each single feature in the database includes:

[0029] Acquire training images and a model to be trained. The training images contain image content of at least two independent parts of the same body part of the object, as well as annotations for each independent part.

[0030] The similarity recognition model is obtained by training the model with training images until it converges.

[0031] A similarity recognition model is used to determine the similarity between a single feature of the target and each single feature in the database.

[0032] In some embodiments, obtaining training images includes:

[0033] Acquire images, and the training images contain the image content of at least two independent parts of the same body part of the object;

[0034] The acquired image is mirrored to obtain a mirrored image;

[0035] Calculate the image similarity between the acquired image and the image acquired after mirroring;

[0036] Images with a similarity greater than a threshold are used as training images.

[0037] In some embodiments, a similarity recognition model is obtained by training training images until convergence, including:

[0038] The image features of the training images are input into the model to be trained to obtain similarity prediction scores between independent parts of the training images;

[0039] When the similarity prediction score matches the label, the image features of the independent parts are magnified to obtain magnified difference features;

[0040] input the amplified difference feature into the to-be-trained model to train the to-be-trained model until the similarity prediction score is inconsistent with the label, perform parameter adjustment processing on the to-be-trained model based on the difference between the similarity prediction score and the label, and obtain the similarity identification model.

[0041] In some embodiments, the obtaining unit is configured to:

[0042] obtain a target image and a target account;

[0043] perform region detection processing on an independent part of a body part of an object in the target image to obtain an independent part region of the target image;

[0044] extract a biometric feature of the independent part region to obtain a target single feature.

[0045] In some embodiments, the database further includes an account address, and the associating unit is configured to:

[0046] obtain a positioning address;

[0047] if the positioning address is the same as the account address, associate the target single feature with the target account.

[0048] Embodiments of the present application also provide an electronic device including a memory storing a plurality of instructions; and a processor loading the instructions from the memory to perform the steps in any of the data processing methods provided by embodiments of the present application.

[0049] Embodiments of the present application also provide a computer-readable storage medium storing a plurality of instructions, the instructions being adapted to be loaded by a processor to perform the steps in any of the data processing methods provided by embodiments of the present application.

[0050] Embodiments of the present application can obtain a target single feature and a target account, the target single feature being a single feature to be associated with the target account, the single feature being one of a plurality of features, the plurality of features being biometric features of at least two independent parts of a same body part of an object; determine an association between the target plurality of features and the target account in a database, the target plurality of features being a plurality of features corresponding to the target single feature, the database including an association relationship between an account and each single feature in the plurality of features; if the target plurality of features and the target account are in a first association in the database, associate the target single feature with the target account; and if the target plurality of features and the target account are in a second association in the database, refuse to associate the target single feature with the target account.

[0051] In the present application, the associated account and single feature can be stored in the database, and whenever a user needs to associate a target single feature and a target account, by determining whether the target multiple feature and the target account exist in the database and their association, it can be determined whether the target multiple feature and the target account can be associated according to the association, the first association case agrees to associate, and the second association case refuses to associate. Thus, the problem of associating multiple accounts with multiple features by the same user in the second association case is solved, and only in the first association case is the association agreed, realizing one person one number, and ensuring the authenticity and security of the account. BRIEF DESCRIPTION OF DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0053] Figure 1a is a scene diagram of the data processing method provided by the embodiments of the present application;

[0054] Figure 1b is a flow diagram of the data processing method provided by the embodiments of the present application;

[0055] Figure 2a is an image acquisition diagram of the data processing method provided by the embodiments of the present application applied in the left and right hand scenario;

[0056] Figure 2b is a model training diagram of the data processing method provided by the embodiments of the present application;

[0057] Figure 2c is a scene diagram of the data processing method provided by the embodiments of the present application;

[0058] Figure 2d is a scene diagram of the data processing method provided by the embodiments of the present application;

[0059] Figure 3 is a structure diagram of the data processing device provided by the embodiments of the present application;

[0060] Figure 4 is a structure diagram of the electronic device provided by the embodiments of the present application. DETAILED DESCRIPTION

[0061] With reference to the drawings of the embodiments of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments of the present application, all the other embodiments obtained by a person of ordinary skill in the art without creative work fall within the scope of the present application.

[0062] The present application provides a data processing method and device, an electronic device, a storage medium and a program product.

[0063] The data processing device can be integrated in an electronic device, which can be a terminal, a server, etc. The terminal can be a mobile phone, a tablet computer, a smart Bluetooth device, a notebook computer, a personal computer (PC), etc. The server can be a single server or a server cluster composed of multiple servers.

[0064] In some embodiments, the data processing device can also be integrated in multiple electronic devices, for example, the data processing device can be integrated in multiple servers, and the multiple servers can be used to implement the data processing method of the present application.

[0065] In some embodiments, the server can also be implemented in the form of a terminal.

[0066] For example, referring to Figure 1a The electronic device can be a server, which can obtain a target single feature and a target account, determine the association of the target multiple features and the target account in the database, associate the target single feature with the target account if the target multiple features and the target account are in a first association in the database, and refuse to associate the target single feature with the target account if the target multiple features and the target account are in a second association in the database.

[0067] In some embodiments, the multiple feature can be a double-hand palmprint, which is composed of two single features, i.e., the double-hand palmprint includes a left-hand palmprint and a right-hand palmprint. For example, embodiments of the present application can associate either or both of the palms of a user's two hands to the same account. If the left-hand palmprint of the user has been associated with account A, when the user is ready to associate the right-hand palmprint with account B, according to the association relationship between the left-hand palmprint of the user and account A in the database, it is determined that the right-hand palmprint of the user and account B are in the second association condition, and the association of the right-hand palmprint of the user and account B is rejected. If the left-hand palmprint of the user has been associated with account A, when the user is ready to associate the right-hand palmprint with account A, according to the association relationship between the left-hand palmprint of the user and account A in the database, it is determined that the right-hand palmprint of the user and account A are in the first association condition, and the association of the right-hand palmprint of the user and account A is agreed, thereby solving the problem of associating multiple accounts by the same user using multiple features, realizing one person one number, and guaranteeing the authenticity and security of the account.

[0068] The following will be described in detail respectively. It should be noted that the serial numbers of the following embodiments are not regarded as a limitation on the preferred order of the embodiments.

[0069] Biometric identification is a technology that uses human body characteristics for identity verification or recognition. It uses the uniqueness of individuals in physiology or behavior to confirm their identity. The development of biometric identification technology benefits from advances in computer vision, machine learning, and artificial intelligence, making recognition systems more intelligent, accurate, and reliable.

[0070] Among them, computer vision (Computer Vision, CV) is a technology that uses computers to replace human eyes to identify, measure, and further process target images. Computer vision technology usually includes image processing, image recognition, image semantic understanding, image retrieval, virtual reality, augmented reality, simultaneous localization and mapping, autonomous driving, intelligent transportation, etc. It also includes common biometric identification technologies such as face recognition and fingerprint recognition. For example, image processing technologies such as image coloring and image edge extraction.

[0071] Machine learning (Machine Learning, ML) is a multi-disciplinary subject that involves probability theory, statistics, approximation theory, convex analysis, algorithm complexity theory, etc. It is a subject that studies how computers simulate or implement human learning behavior to acquire new knowledge or skills and reorganize existing knowledge structure to continuously improve their performance. Machine learning is the core of artificial intelligence and the fundamental approach to making computers intelligent. It is applied in various fields of artificial intelligence. Machine learning and deep learning usually include artificial neural networks, belief networks, reinforcement learning, transfer learning, inductive learning, and supervised learning.

[0072] In the embodiment, a data processing method based on biometric technology involving artificial intelligence is provided, as shown in the figure, the specific process of the data processing method can be as follows: Figure 1b

[0073] 110, obtaining a target single feature and a target account.

[0074] Biometric technology can be divided into unique features and multiple features according to the nature of the features.

[0075] Unique features refer to the biometric features of the unique body parts of the object, usually the features of some specific parts or organs of the human body, such as facial features.

[0076] Multiple features refer to the biometric features of at least two independent parts of the same body part of the object, usually including the features of different parts of the same part, such as the irises of the left and right eyes, the palm prints of the left and right hands, and the fingerprints of the ten fingers.

[0077] Multiple features are composed of at least one single feature, for example, the irises of both eyes are a multiple feature composed of two single features, the left eye iris and the right eye iris; for example, the palm prints of both hands are a multiple feature composed of two single features, the left hand palm print and the right hand palm print; for example, the fingerprints of ten fingers are a multiple feature composed of ten single features, i.e. the fingerprints of ten fingers.

[0078] The target single feature is a single feature to be associated with the target account. In some embodiments, the target user wants to bind his left hand palm print to his social media account, so the left hand palm print and the social media account can be obtained from the user terminal of the target user.

[0079] In some embodiments, the user can collect his own left hand photo by taking a photo, and the device proposed in the embodiment of the present application extracts the left hand palm print from it, therefore, step 110 includes the following steps:

[0080] Obtaining a target image and a target account;

[0081] Performing region detection processing on the independent parts of the body part of the object in the target image to obtain the independent part region of the target image;

[0082] Extracting the biometric features of the independent part region to obtain the target single feature.

[0083] ​For example, in some embodiments, the user can use a mobile terminal to open an application, the application opens a camera, guides the user to place the left hand in a proper position, and takes a photo of the left hand, after the user finishes taking the photo, the application uploads the photo of the palm to a server, the server performs a region detection process on the left hand in the photo of the palm, obtains a left hand region in the photo of the palm, extracts a palmprint feature of the left hand region in the photo of the palm, and obtains a left palmprint of the target user.

[0084] In the region detection process, a region in which an independent part appears in a target image can be identified, for example, a left hand region in a photo of a palm, so as to remove the influence of other image contents.

[0085] In some embodiments, the target image can include one or more of a normal image, an infrared image, a depth image, and the like.

[0086] In some embodiments, when the target image is acquired, in order to ensure that the object body part in the acquired target image is a real existence and a live body, rather than a static image or a photocopy, and the like, a live body verification can be performed when the target image is collected, the target image passing the live body verification is retained, and the target image failing the live body verification is rejected.

[0087] In the live body verification, a method can include texture analysis, shape change detection, thermal infrared detection, depth information analysis, dynamic interaction detection, and the like, so as to prevent the use of a photo or other fraudulent means for association.

[0088] In some embodiments, when the target image is acquired, in order to ensure that the image quality of the acquired target image meets a certain standard, and to exclude the interference of low-quality images, a quality verification can be performed when the target image is collected, the target image passing the quality verification is retained, and the target image failing the quality verification is rejected.

[0089] In the quality verification, a method can include image sharpness evaluation, contrast evaluation, texture evaluation, and the like.

[0090] In some embodiments, if the target user submits a target image containing left and right hands, a symmetry verification can be performed on the target image, the target image passing the symmetry verification is retained, and the target image failing the symmetry verification is rejected, so as to avoid data source dislocation and confusion, and to ensure that the left and right hands of the target user are successfully recorded.

[0091] In the symmetry verification, a mirror image processing is performed on the target image, a similarity calculation is performed on the target image and the target image after the mirror image processing, so as to determine whether the target image before and after the mirror image processing is symmetrical, and if so, it means that the left and right hands of the target user are successfully recorded.

[0092] 120, determine the association of the target multi-feature and the target account in the database.

[0093] The target multi-feature is a multi-feature corresponding to a target single feature, and the database includes an association relationship between an account and each single feature in the multi-feature.

[0094] The association in the database has multiple types, for example, referring to Table 1, in the scenario where the multi-feature is composed of two single features, for example, the dual palm print scenario, if the target user wants to associate the left palm print a with the account A, and there are 10 subcases in the database:

[0095] Number Left-hand palm print Right-hand palm print Account number 1 a a’ A 2 a’ A 3 a A 4 b b’ A 5 b A 6 b’ A 7 a a’ B 8 a B 9 a’ B 10

[0096] Table 1

[0097] 1. The database has an association relationship between account A and left palm print a and right palm print a’;

[0098] 2. The database has an association relationship between account A and right palm print a’;

[0099] 3. The database has an association relationship between account A and left palm print a;

[0100] 4. The database has an association relationship between account A and left palm print b and right palm print b’;

[0101] 5. The database has an association relationship between account A and right palm print b’;

[0102] 6. The database has an association relationship between account A and left palm print b;

[0103] 7. The database has an association relationship between account B and left palm print a and right palm print a’;

[0104] 8. The database has an association relationship between account B and right palm print a’;

[0105] 9. The database has an association relationship between account B and left palm print a;

[0106] 10. The database does not have an association relationship between account A and left palm print a.

[0107] The subcases 1, 2, 3, and 10 can be collectively referred to as the first association, which means that the database has an association relationship between the target multi-feature and the target account, or the database does not have the target multi-feature and the target account.

[0108] For example, the target user wants to associate the left palm print a with the account A. Since the sub-case 1 and 3, the database already has the association between the left palm print a and the account A, so no extra action is needed to achieve the step 130.

[0109] For example, the target user wants to associate the left palm print a with the account A. Since the sub-case 2 and 10, the database does not have the association between the left palm print a and the account A, but the left palm print a can be associated with the account A, so the step 130 is performed.

[0110] The sub-cases 4-9 can be collectively referred to as the second association case. The second association case refers to the association between the target multi-feature and other accounts in the database, i.e., the left palm print a of the target user has been associated with the account of other users. Or there is an association between other multi-features and the target account, i.e., the target account A of the target user has been associated with the palm print of other users, so the left palm print a cannot be associated with the account A. The step 140 is performed to reject the association of the left palm print a with the account A.

[0111] Therefore, in some embodiments, the step 120 includes the following steps:

[0112] Respectively detecting the association relationship of the target multi-feature in the database, and the association relationship of the target account;

[0113] If the database has an association relationship between the target multi-feature and the target account, or the database does not have the target multi-feature and the target account, it is determined that the target multi-feature and the target account are in the first association case in the database.

[0114] If the database has an association relationship between the target multi-feature and other accounts, or has an association relationship between other multi-features and the target account, it is determined that the target multi-feature and the target account are in the second association case in the database. The other accounts are other accounts except the target account, and the other multi-features are other multi-features except the target multi-feature.

[0115] In some embodiments, a quick search process can be set to quickly screen the sub-case 10, i.e., the step 120 further includes:

[0116] Performing a quick search process on the target single feature in the database, thereby obtaining a quick search result of the target single feature in the database;

[0117] If the quick search result indicates that the database does not have the target single feature, it is determined that the target multi-feature and the target account are in the first association case in the database.

[0118] In some embodiments, a Bloom Filter can be utilized to enable fast data lookup in the database.

[0119] In some embodiments, the existence of the target single feature in the database can be confirmed by calculating the similarity, and thus the association of the target multiple features in the database and the association of the target account are detected respectively, including the following steps:

[0120] determining the similarity between the target single feature and each single feature in the database;

[0121] determining similar features to the target single feature in the database according to the similarity;

[0122] determining the account associated with the similar features;

[0123] determining the single features associated with the target account in the database.

[0124] In some embodiments, the similarity calculation can be implemented by a neural network model. Thus, the determination of the similarity between the target single feature and each single feature in the database includes:

[0125] obtaining training images and a model to be trained, the training images containing image content of at least two independent parts of the same body part of an object, and annotations for each independent part;

[0126] training the model to be trained with the training images until convergence to obtain a similarity recognition model;

[0127] determining the similarity between the target single feature and each single feature in the database using the similarity recognition model.

[0128] For example, in some embodiments, the training images can be images of both hands, and the training images can be annotated with the regions of the left and right hands, the positions and shapes of the palm prints, the regions annotated with the left and right hands, and the similarity between the left and right hands, etc.

[0129] In some embodiments, the model to be trained can include any neural network model such as a convolutional neural network (CNN), an autoencoder (Autoencoder), etc.

[0130] In some embodiments, backpropagation and optimization algorithms can be used to enable the model to learn the similarity features between the left and right hands and use them to distinguish the left and right hands of different people.

[0131] In some embodiments, the above-mentioned living body verification, quality verification, symmetry verification and the like mechanisms can be set to preprocess the training images. For example, a mirror image verification can be set to ensure that the left and right palm images are collected in pairs, avoid misplacement and confusion of the data source, and ensure that the left and right palm images of the same person are collected. Therefore, the step of obtaining the training images includes the following steps:

[0132] obtaining the collected images, the training images containing image content of at least two independent parts in the same body part of the subject;

[0133] performing mirror image processing on the collected images to obtain mirror image collected images;

[0134] calculating the image similarity between the collected images and the mirror image collected images;

[0135] taking the collected images with the image similarity greater than a threshold value as the training images.

[0136] In some embodiments, a difference amplification mechanism can be set for the similarity recognition model to amplify the differences between the left and right hands of different people and shrink the differences between the left and right hands of the same person to a smaller extent, so as to improve the accuracy and robustness of the model. Therefore, the step of training the to-be-trained model with the training images until convergence to obtain the similarity recognition model includes the following steps:

[0137] inputting the image features of the training images into the to-be-trained model to obtain a similarity prediction score between the independent parts in the training images;

[0138] when the similarity prediction score is consistent with the label, performing difference amplification processing on the image features of the independent parts to obtain amplified difference features;

[0139] inputting the amplified difference features into the to-be-trained model to train the to-be-trained model until the similarity prediction score is inconsistent with the label, and performing parameter adjustment processing on the to-be-trained model based on the difference between the similarity prediction score and the label to obtain the similarity recognition model.

[0140] In some embodiments, the difference amplification processing can include a sharpening algorithm, a statistical difference amplification algorithm, a non-local mean algorithm and the like.

[0141] 130、if the target multiple features and the target account are in the first association condition in the database, then associate the target single feature with the target account.

[0142] In some embodiments, a risk control mechanism based on geographic positioning can be set to ensure that the user binds his / her own target account using his / her own target single feature. Therefore, the database further includes an account address, and the step of associating the target single feature with the target account includes the following steps:

[0143] obtaining a positioning address;

[0144] If the positioning address is the same as the account address, the target single feature is associated with the target account;

[0145] Otherwise, the target single feature is rejected from being associated with the target account.

[0146] 140、If the target multiple feature and the target account are in the second association condition in the database, the target single feature is rejected from being associated with the target account.

[0147] In some embodiments, the target user can also be prompted for rejection.

[0148] As can be seen from the above, the embodiments of the present application can obtain a target single feature and a target account, the target single feature being a single feature to be associated with the target account, the single feature being one of multiple features in an object, the multiple features being biological recognition features of at least two independent parts of the same body part; determine the association condition of the target multiple feature and the target account in the database, the target multiple feature being a multiple feature corresponding to the target single feature, the database including an association relationship between an account and each single feature in the multiple features; if the target multiple feature and the target account are in the first association condition in the database, the target single feature is associated with the target account; if the target multiple feature and the target account are in the second association condition in the database, the target single feature is rejected from being associated with the target account. Thus, the present application solves the problem of associating multiple accounts with the same user using multiple features, realizes one person one account, and ensures the authenticity and security of the account.

[0149] According to the method described in the above embodiments, the following will be further described in detail.

[0150] In this embodiment, the method of the present application will be described in detail by taking left and right palm prints as examples.

[0151] (I) Image acquisition

[0152] In some embodiments, the user can be prompted to show the left and right hands in front of the camera respectively, and the acquisition of the double-hand pictures can be realized through quality verification, live body verification and symmetry verification.

[0153] In some embodiments, the RGB image of the double hands and the infrared image of the palm vein can be acquired at the same time to obtain more comprehensive palm feature information.

[0154] For example, refer to Figure 2arespectively, to determine whether the palm is a real live body, prevent image acquisition by using photos or other fraudulent means, and improve the security and reliability of the system; then, the left and right hand images are respectively subjected to quality checking to ensure that the image clarity, contrast, and texture indicators meet certain standards, and to exclude interference from low-quality images; finally, the left and right hand images are subjected to symmetry checking, the left hand image is mirror-processed to obtain a mirror-processed left hand image, and the right hand image is mirror-processed to obtain a mirror-processed right hand image, and then the left hand image and the mirror-processed right hand image are subjected to symmetry checking, and the right hand image and the mirror-processed left hand image are subjected to symmetry checking, to avoid data source misplacement and confusion and to ensure that the left and right hand palm images of the same person are collected.

[0155] (ii) Training of the similarity recognition model.

[0156] In some embodiments, the similarity recognition model can be trained in the training stage to learn paired left and right hand samples, so as to mine the similarity of the left and right hands, so that the similarity recognition model can distinguish between different people or the left and right hands of the same person.

[0157] The training sample can be an acquisition image, in which the boundary box of the hand, the position and shape of the palm print, and the similarity of the two hands can be labeled.

[0158] In some embodiments, the similarity recognition model is provided with a difference amplification mechanism, for example, referring to Figure 2b For training the pre-trained model with the training sample, the trained model can predict the similarity prediction score of the left and right hands in the training sample. If the similarity prediction score is consistent with the true score in the label, the image features of the training sample are subjected to difference amplification processing, and the difference amplified image features are re-input into the trained model until the similarity prediction score is inconsistent with the true score in the label. The trained model is adjusted according to the difference between the similarity prediction score and the true score in the label until the trained model converges.

[0159] The difference amplification processing can amplify the difference between the left and right hand palm images of different people, so that the model can better distinguish the other hand that is not the person's own, improve the ability to distinguish the left and right hands of different people, and thus provide more accurate and reliable results for the authentication and recognition tasks of the palm image.

[0160] (iii) Application of the similarity recognition model.

[0161] In some embodiments, the similarity recognition model can quickly and accurately determine whether the input left and right hand palm images are the person's own palm according to the input left and right hand palm images. By distinguishing the similarity of the left and right hands and amplifying the difference, the module can help the judgment module to more effectively distinguish the nature of the left and right hands, thereby improving the accuracy of the judgment.

[0162] In some embodiments, due to the difference amplification mechanism set by the similarity recognition model, the left and right hands of different people can be more obviously distinguished, the accuracy of judgment can be improved, and the probability of false judgment of non-personal palm can be reduced.

[0163] Reference Figure 2c If the left and right hand pictures or the target account of the target user are determined to be absent in the database through the Bloom filter, it can be directly determined that the left and right hand pictures of the user are in the first association condition with the target account in the database, and the left and right hand pictures of the user can be directly associated with the target account; if it is determined to exist in the database, the similarity discrimination model can be used for further accurate determination.

[0164] Through the similarity recognition model based on deep learning, it can quickly and accurately determine whether the input palm image is the palm of the person, and through the amplification of the difference and the analysis of the similarity characteristics, the module can provide reliable determination results, and provide efficient and accurate support for the authentication and recognition tasks of the system.

[0165] (Four) Intercepting mechanism.

[0166] The intercepting mechanism proposed in the present application ensures that one person has one account, thereby improving the security of the account, that is, agreeing to association in the first association condition and refusing association in the second association condition.

[0167] Reference Figure 2d An intercepting mechanism based on geographic positioning is also proposed, which is used to judge the risk degree of the account association request, so as to avoid associating the account of the person with the palmprint of the non-person, or associating the account of the non-person with the palmprint of the person.

[0168] In some embodiments, in order to improve the accuracy and comprehensiveness of the interception, a risk control mechanism such as behavior analysis, device fingerprint identification, IP address verification, etc. can be combined to judge the risk degree of the account association request, so as to more accurately judge whether to refuse association.

[0169] As can be seen from the above, the embodiments of the present application can solve the problem of associating multiple accounts with multiple features of the same user, realize one person one account, prevent malicious users from associating accounts, and protect the authenticity and security of the account.

[0170] It can be understood that in the specific embodiments of the present application, biological recognition features, collected images and other related data are involved, and when the following embodiments of the present application are applied to specific products or technologies, permission or consent is required, and the collection, use and processing of related data need to comply with relevant laws, regulations and standards of relevant countries and regions.

[0171] To better implement the above methods, this application also provides a data processing device, which can be integrated into an electronic device, such as a terminal or server. The terminal can be a mobile phone, tablet computer, smart Bluetooth device, laptop computer, personal computer, etc.; the server can be a single server or a server cluster composed of multiple servers.

[0172] For example, in this embodiment, the method of this application embodiment will be described in detail by taking the data processing device specifically integrated into the server as an example.

[0173] For example, such as Figure 3 As shown, the data processing device may include an acquisition unit 301, a determination unit 302, an association unit 303, and a rejection unit 304, as follows:

[0174] (a) Acquisition Unit 301.

[0175] The acquisition unit 301 is used to acquire the target single feature and the target account. The target single feature is a single feature to be associated with the target account. The single feature is one of the multiple features. The multiple features are biometric features of at least two independent parts of the same body part of the object.

[0176] In some embodiments, the acquisition unit 301 is configured to:

[0177] Obtain the target image and target account;

[0178] Region detection processing is performed on independent parts of the body parts of the object in the target image to obtain independent part regions of the target image;

[0179] Biometric features of independent regions are extracted to obtain single-target features.

[0180] (ii) Determine unit 302.

[0181] The determining unit 302 is used to determine the association between the target multiple features and the target account in the database. The target multiple features are multiple features corresponding to the target single feature. The database includes the association between the account and each single feature in the multiple features.

[0182] In some embodiments, the determining unit 302 includes:

[0183] The detection subunit is used to detect the association relationships of multiple features of the target in the database, as well as the association relationships of the target accounts;

[0184] The first subunit is configured to determine that the target multi-feature and the target account are in a first association condition in the database if there is an association relationship between the target multi-feature and the target account in the database, or there is no target multi-feature and target account in the database.

[0185] The second subunit is configured to determine that the target multi-feature and the target account are in a second association condition in the database if there is an association relationship between the target multi-feature and other accounts in the database, or there is an association relationship between other multi-features and the target account, the other accounts being accounts other than the target account, and the other multi-features being multi-features other than the target multi-feature.

[0186] In some embodiments, the determination unit 302 is further configured to:

[0187] perform a quick search on the target single feature in the database to obtain a quick search result of the target single feature in the database.

[0188] If the quick search result indicates that the target single feature does not exist in the database, it is determined that the target multi-feature and the target account are in the first association condition in the database.

[0189] In some embodiments, the detection subunit is configured to:

[0190] determine the similarity between the target single feature and each single feature in the database;

[0191] determine a similar feature similar to the target single feature in the database according to the similarity;

[0192] determine an account associated with the similar feature;

[0193] determine a single feature associated with the target account in the database.

[0194] In some embodiments, determining the similarity between the target single feature and each single feature in the database comprises:

[0195] obtaining a training image and a to-be-trained model, the training image containing image content of at least two independent parts of the same body part of an object, and a label for each independent part;

[0196] training the to-be-trained model with the training image until convergence to obtain a similarity recognition model;

[0197] determining the similarity between the target single feature and each single feature in the database using the similarity recognition model.

[0198] In some embodiments, obtaining the training image comprises:

[0199] acquire the collected image, the training image contains image content of at least two independent parts of the same body part of the object;

[0200] mirror the collected image to obtain a mirrored collected image;

[0201] calculate the image similarity between the collected image and the mirrored collected image;

[0202] take the collected image with the image similarity greater than the threshold as the training image.

[0203] In some embodiments, the training image is used to train the to-be-trained model until convergence, to obtain the similarity identification model, including:

[0204] input the image features of the training image into the to-be-trained model to obtain the similarity prediction score between the independent parts in the training image;

[0205] when the similarity prediction score is consistent with the label, difference amplification processing is performed on the image features of the independent parts to obtain amplified difference features;

[0206] the amplified difference features are input into the to-be-trained model to train the to-be-trained model until the similarity prediction score is inconsistent with the label, and the to-be-trained model is adjusted based on the difference between the similarity prediction score and the label to obtain the similarity identification model.

[0207] (three) the association unit 303.

[0208] The association unit 303 is configured to associate the target single feature with the target account if the target multiple features and the target account are in the first association condition in the database.

[0209] In some embodiments, the database further includes an account address, and the association unit 303 is configured to:

[0210] acquire the positioning address;

[0211] if the positioning address is the same as the account address, associate the target single feature with the target account.

[0212] (four) the rejection unit 304.

[0213] The rejection unit 304 is configured to reject the association of the target single feature with the target account if the target multiple features and the target account are in the second association condition in the database.

[0214] In specific implementation, each of the above units can be implemented as an independent entity, or can be combined as the same or several entities, and the specific implementation of each of the above units can be referred to the method embodiments above, which will not be described here.

[0215] In this application embodiment, the terms "module" or "unit" refer to a computer program or part of a computer program that has a predetermined function and works with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit.

[0216] As described above, the data processing device in this embodiment acquires a target single feature and a target account through an acquisition unit. The target single feature is a single feature to be associated with the target account. The single feature is one of the multiple features constituting multiple features. Multiple features are biometric features of at least two independent parts of the same body part of an object. The determination unit determines the association between the target multiple features and the target account in the database. The target multiple features are multiple features corresponding to the target single feature. The database includes the association relationship between the account and each single feature in the multiple features. If the target multiple features and the target account are in a first association state in the database, the association unit associates the target single feature with the target account. If the target multiple features and the target account are in a second association state in the database, the rejection unit rejects the association between the target single feature and the target account. Therefore, this embodiment can solve the problem of the same user associating multiple accounts using multiple features, achieving one account per person and ensuring the authenticity and security of the account.

[0217] This application also provides an electronic device, which can be a terminal, a server, or other similar device. The terminal can be a mobile phone, tablet computer, smart Bluetooth device, laptop computer, personal computer, etc.; the server can be a single server or a server cluster composed of multiple servers, etc.

[0218] In some embodiments, the data processing apparatus may also be integrated into multiple electronic devices, such as multiple servers, with the data processing method of this application being implemented by the multiple servers.

[0219] In this embodiment, a server will be used as an example for detailed description. For example, ... Figure 4 As shown, it illustrates a structural schematic diagram of the electronic device involved in the embodiments of this application, specifically:

[0220] The electronic device may include components such as a processor 401 with one or more processing cores, a memory 402 with one or more computer-readable storage media, a power supply 403, an input module 404, and a communication module 405. Those skilled in the art will understand that... Figure 4The electronic device structure shown in the figures is not intended to limit the electronic device, which can include more or fewer components than shown, or have components combined together or arranged differently. Among others:

[0221] The processor 401 is the control center of the electronic device, which connects various parts of the electronic device through various interfaces and lines, executes various functions of the electronic device and processes data by running or executing software programs and / or modules stored in the memory 402 and calling data stored in the memory 402, thereby overall detecting the electronic device. In some embodiments, the processor 401 can include one or more processing cores; in some embodiments, the processor 401 can integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application program, etc., and the modem processor mainly processes wireless communication. It can be understood that the above-mentioned modem processor can also not be integrated into the processor 401.

[0222] The memory 402 can be used to store software programs and modules, and the processor 401 executes various function applications and data processing by running the software programs and modules stored in the memory 402. The memory 402 can mainly include a program storage area and a data storage area, wherein the program storage area can store the operating system, at least one application program required by the function (such as sound playing function, image playing function, etc.), etc.; the data storage area can store data created according to the use of the electronic device, etc. In addition, the memory 402 can include a high-speed random access memory, and can also include a non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other volatile solid-state memory device. Accordingly, the memory 402 can also include a memory controller to provide the processor 401 with access to the memory 402.

[0223] The electronic device also includes a power supply 403 for supplying power to various components, and in some embodiments, the power supply 403 can be logically connected to the processor 401 through a power management system, so as to realize the functions of managing charging, discharging, and power consumption management, etc. through the power management system. The power supply 403 can also include one or more direct current or alternating current power supplies, recharging systems, power failure detection circuits, power converters or inverters, power state indicators, etc. any components.

[0224] The electronic device can also include an input module 404, which can be used to receive input digital or character information, and generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function control.

[0225] The electronic device can also include a communication module 405, which in some embodiments can include a wireless module, through which the electronic device can perform short-range wireless transmission, thereby providing the user with wireless broadband Internet access. For example, the communication module 405 can be used to help the user send and receive emails, browse web pages, and access streaming media, etc.

[0226] Although not shown, the electronic device can also include a display unit, etc., which will not be described here. In particular, in the present embodiment, the processor 401 in the electronic device will load the executable file corresponding to the process of one or more application programs into the memory 402 according to the following instructions, and run the application program stored in the memory 402 by the processor 401, thereby realizing various functions, such as:

[0227] Obtaining a target single feature and a target account, the target single feature being a single feature to be associated with the target account, the single feature being one of a plurality of features, the plurality of features being biological recognition features of at least two independent parts of the same body part of an object;

[0228] Determining an association between the target multi-feature and the target account in the database, the target multi-feature being a multi-feature corresponding to the target single feature, the database including an association relationship between the account and each single feature in the multi-feature;

[0229] If the target multi-feature and the target account are in the first association in the database, then associate the target single feature with the target account;

[0230] If the target multi-feature and the target account are in the second association in the database, then refuse to associate the target single feature with the target account.

[0231] The specific implementation of each operation can refer to the previous embodiments, which will not be described here.

[0232] As can be seen from the above, the embodiments of the present application can solve the problem of associating multiple accounts with the same user using multiple features, realize one account per user, and ensure the authenticity and security of the account.

[0233] Those of ordinary skill in the art can understand that all or part of the steps of the various methods of the above embodiments can be completed by instructions, or by instructions controlling related hardware, which can be stored in a computer readable storage medium and loaded and executed by a processor.

[0234] To this end, the embodiments of the present application provide a computer readable storage medium, which stores a plurality of instructions, which can be loaded by a processor to execute the steps in any data processing method provided by the embodiments of the present application. For example, the instructions can perform the following steps:

[0235] obtaining a target single feature and a target account, the target single feature being a single feature to be associated with the target account, the single feature being one of a plurality of features, the plurality of features being biometric features of at least two independent parts of a same body part of the object;

[0236] determining an association between the target plurality of features and the target account in a database, the target plurality of features being a plurality of features corresponding to the target single feature, the database including an association between the account and each single feature of the plurality of features;

[0237] if the target plurality of features and the target account are in a first association in the database, then associating the target single feature with the target account;

[0238] if the target plurality of features and the target account are in a second association in the database, then rejecting the association of the target single feature with the target account.

[0239] The storage medium can include a read-only memory (ROM), a random access memory (RAM), a magnetic disk, an optical disk, or the like.

[0240] According to an aspect of the present application, a computer program product or computer program is provided, which includes computer instructions stored in a computer readable storage medium. A processor of a computer device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions to cause the computer device to perform the method provided in any of the various optional implementation manners of the account binding aspect or the account association aspect provided in the above embodiments.

[0241] Due to the instructions stored in the storage medium, the steps of any of the data processing methods provided in the embodiments of the present application can be performed, and thus the beneficial effects of any of the data processing methods provided in the embodiments of the present application can be achieved. Details are described in the above embodiments, and thus will not be described here again.

[0242] The above describes in detail a data processing method, device, electronic device and computer readable storage medium provided in the embodiments of the present application. The principles and implementation manners of the present application are described by applying specific examples in this paper. The above embodiment descriptions are only used to help understand the method of the present application and its core idea. Meanwhile, for those skilled in the art, according to the idea of the present application, the specific implementation manners and application ranges will be changed. In summary, the content of the specification should not be understood as a limitation of the present application.

Claims

1. A data processing method, characterized in that, include: Obtain a single target feature and a target account. The single target feature is a single feature to be associated with the target account. The single feature is one of multiple features. The multiple features are biometric features of at least two independent parts of the same body part of the object. Determine the association between the target's multiple features and the target account in the database. The target's multiple features are the multiple features corresponding to the target's single feature. The database includes the association between the account and each single feature in the multiple features. If the target's multiple features and the target account are in a first association situation in the database, then the target's single feature is associated with the target account; If the target's multiple features and the target account are in a second association situation in the database, then the association between the target's single feature and the target account is rejected.

2. The data processing method as described in claim 1, characterized in that, The determination of the target's multiple features and the association between the target account and the database includes: The association relationships of the target's multiple features in the database and the association relationships of the target's accounts are detected respectively. If the database contains an association between the target multiple features and the target account, or if the database does not contain the target multiple features and the target account, then the target multiple features and the target account are determined to be in the first association situation in the database. If the database contains associations between the target's multiple features and other accounts, or between other multiple features and the target account, then the target's multiple features and the target account are determined to be in a second association state in the database. The other accounts are accounts other than the target account, and the other multiple features are multiple features other than the target's multiple features.

3. The data processing method as described in claim 2, characterized in that, The determination of the target's multiple features and the association between the target account and the database also includes: A fast search process is performed on the target single feature in the database to obtain the fast search result of the target single feature in the database; If the quick search result indicates that the target single feature does not exist in the database, then the target multiple features and the target account are determined to be in the first association situation in the database.

4. The data processing method as described in claim 2, characterized in that, The step of detecting the association relationships of the target's multiple features in the database and the association relationships of the target account includes: Determine the similarity between the target single feature and each single feature in the database; Based on the similarity, similar features that are similar to the target single feature are determined in the database; Identify the accounts associated with the aforementioned similar characteristics; A single characteristic associated with the target account is identified in the database.

5. The data processing method as described in claim 4, characterized in that, Determining the similarity between the target single feature and each single feature in the database includes: Acquire training images and a model to be trained. The training images contain image content of at least two independent parts of the same body part of the object, as well as annotations for each independent part. The training images are used to train the model to be trained until convergence, thus obtaining a similarity recognition model; The similarity recognition model is used to determine the similarity between the target single feature and each single feature in the database.

6. The data processing method as described in claim 5, characterized in that, The acquisition of training images includes: Acquire images, wherein the training images contain image content of at least two independent parts of the same body part of the object; The acquired image is mirrored to obtain a mirrored acquired image; Calculate the image similarity between the acquired image and the image acquired after mirroring; Images with a similarity greater than a threshold are used as training images.

7. The data processing method as described in claim 5, characterized in that, The step of training the model to be trained using the training images until convergence to obtain a similarity recognition model includes: The image features of the training image are input into the model to be trained to obtain similarity prediction scores between independent parts in the training image; When the similarity prediction score matches the label, the image features of the independent part are subjected to difference amplification processing to obtain amplified difference features; The amplified difference features are input into the model to be trained to train the model until the similarity prediction score is inconsistent with the label. Based on the difference between the similarity prediction score and the label, the parameters of the model to be trained are tuned to obtain the similarity recognition model.

8. The data processing method as described in claim 1, characterized in that, The acquisition of the target single feature and the target account includes: Obtain the target image and target account; Perform region detection processing on independent parts of the body parts of the object in the target image to obtain independent part regions of the target image; Biometric features of the independent partial regions are extracted to obtain the target single feature.

9. The data processing method as described in claim 1, characterized in that, The database also includes account addresses, and associating the target single feature with the target account includes: Get the location address; If the location address is the same as the account address, then the target single feature is associated with the target account.

10. A data processing apparatus, characterized in that, include: The acquisition unit is used to acquire a target single feature and a target account. The target single feature is a single feature to be associated with the target account. The single feature is one of multiple features. The multiple features are biometric features of at least two independent parts of the same body part of the object. A determining unit is used to determine the association between target multiple features and target account in the database, wherein the target multiple features are multiple features corresponding to the target single feature, and the database includes the association relationship between the account and each single feature in the multiple features; The association unit is used to associate the target single feature with the target account if the target multiple features and the target account are in a first association situation in the database; The rejection unit is used to refuse to associate the target single feature with the target account if the target multiple features and the target account are in a second association situation in the database.

11. The data processing method as described in claim 10, characterized in that, The determining unit includes: The detection subunit is used to detect the association relationships of the target's multiple features in the database, as well as the association relationships of the target account; The first subunit is used to determine that the target multiple features and the target account are in a first association situation in the database if there is an association relationship between the target multiple features and the target account in the database, or if there is no association relationship between the target multiple features and the target account in the database. The second subunit is used to determine that the target multiple features and the target account are in a second association state in the database if there is an association relationship between the target multiple features and other accounts in the database, or an association relationship between other multiple features and the target account. The other accounts are accounts other than the target account, and the other multiple features are multiple features other than the target multiple features.

12. The data processing method as described in claim 11, characterized in that, The determining unit is further configured to: A fast search process is performed on the target single feature in the database to obtain the fast search result of the target single feature in the database; If the quick search result indicates that the target single feature does not exist in the database, then the target multiple features and the target account are determined to be in the first association situation in the database.

13. An electronic device, characterized in that, The method includes a processor and a memory, the memory storing multiple instructions; the processor loads instructions from the memory to perform the steps of the data processing method as described in any one of claims 1 to 9.

14. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a plurality of instructions adapted for loading by a processor to perform the steps of the data processing method according to any one of claims 1 to 9.

15. A computer program product, characterized in that, It includes a computer program / instruction that, when executed by a processor, implements the steps of the data processing method according to any one of claims 1 to 9.

Images