Biological information identification method and device, equipment, storage medium and program product

By processing the depth of field and background area of ​​the recognized image, the target blur image is generated, which solves the problem of private information leakage in biological information identification and significantly improves privacy security.

CN119068508APending Publication Date: 2024-12-03TENCENT TECHNOLOGY (SHENZHEN) CO LTD
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202310636003.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2023-05-31
Publication Date
2024-12-03

AI Technical Summary

Technical Problem

In the prior art, when bioinformatics is identified, the direct processing of the image to be identified leads to the leakage of biological privacy information, which is poor in privacy and security.

Method used

By performing depth of field blurring on the image to be recognized, the background area is determined and blurred, a target blur image is generated, and biological information is recognized based on the image.

Benefits of technology

It significantly improves the privacy and security of bioinformatics identification, and effectively protects the privacy of bioinformatics through the blur effect of overlapping areas of two fuzzy processes.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119068508A_ABST
    Figure CN119068508A_ABST
Patent Text Reader

Abstract

The invention provides a biological information identification method and device, equipment, a storage medium and a program product. The method comprises the steps of obtaining a to-be-recognized image, wherein a foreground of the to-be-recognized image comprises biological information; performing depth-of-field blurring on the to-be-recognized image to obtain a depth-of-field blurred image corresponding to the to Determining a background area corresponding to the background of the depth-of-field blurred image; performing blurring processing on a background region in the depth-of-field blurred image to obtain a target blurred image corresponding to the to-be-recognized image; and based on the target blurred image, identifying the biological information to obtain a corresponding identification result. According to the invention, the privacy security of biological information identification can be effectively improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a method, device, equipment, storage medium and program product for biometric information recognition. Background Art

[0002] Artificial Intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. In other words, artificial intelligence is a comprehensive technology in computer science that attempts to understand the essence of intelligence and produce a new intelligent machine that can respond in a similar way to human intelligence. Artificial intelligence is to study the design principles and implementation methods of various intelligent machines so that machines have the functions of perception, reasoning and decision-making.

[0003] In the related art, biometric information recognition is usually performed directly on the image to be recognized. Since the image to be recognized contains a large amount of biological privacy information, if the biometric information is directly recognized on the image to be recognized, the biological privacy information will be leaked, resulting in poor privacy security of biometric information recognition. Summary of the invention

[0004] The embodiments of the present application provide a biometric information recognition method, device, electronic device, computer-readable storage medium and computer program product, which can effectively improve the privacy security of biometric information recognition.

[0005] The technical solution of the embodiment of the present application is implemented as follows:

[0006] The present application provides a method for biometric information recognition, including:

[0007] Acquire an image to be identified, wherein the foreground of the image to be identified includes biological information;

[0008] Performing depth-of-field blurring on the image to be identified to obtain a depth-of-field blurred image corresponding to the image to be identified;

[0009] Determining a background area corresponding to the background of the depth-of-field blurred image;

[0010] Performing blur processing on the background area in the depth-of-field blurred image to obtain a target blurred image corresponding to the image to be identified;

[0011] Based on the target blurred image, the biological information is recognized to obtain a corresponding recognition result.

[0012] The present application provides a biometric information recognition device, including:

[0013] An acquisition module, used for acquiring an image to be identified, wherein the foreground of the image to be identified includes biological information;

[0014] A depth-of-field blur module, used to perform depth-of-field blur on the image to be identified, to obtain a depth-of-field blurred image corresponding to the image to be identified;

[0015] A determination module, used to determine a background area corresponding to the background of the depth-of-field blurred image;

[0016] A blur module, used for blurring the background area in the depth-of-field blurred image to obtain a target blurred image corresponding to the image to be identified;

[0017] The recognition module is used to recognize the biological information based on the target blurred image to obtain a corresponding recognition result.

[0018] In the above scheme, the above-mentioned depth of field blur module is also used to obtain the target pixel point in the image to be identified, the target pixel point is the pixel point in the image to be identified that coincides with the focus of the image acquisition device that acquires the image to be identified; obtain the depth of field distance, the depth of field distance is less than the maximum value of the distance between the target pixel point and each other pixel point in the image to be identified, the other pixel points are the pixel points in the image to be identified other than the target pixel point; for each of the other pixel points, when the distance between the other pixel point and the target pixel point is greater than the depth of field distance, the other pixel point is determined as a depth of field pixel point; and blur each of the depth of field pixel points in the image to be identified respectively to obtain the depth of field blurred image.

[0019] In the above scheme, the above blur module is also used to obtain the depth value of each pixel point in the image to be identified, and the depth threshold; for each pixel point, when the depth value of the pixel point is greater than the depth threshold, the pixel point is determined as a depth of field pixel point; each depth of field pixel point in the image to be identified is blurred separately to obtain the depth of field blurred image.

[0020] In the above scheme, the above determination module is also used to obtain a target prediction model and call the target prediction model to perform background prediction on each pixel in the depth of field blurred image to obtain the probability that each pixel is a background pixel, and the background pixel is a pixel in the background of the depth of field blurred image; each pixel whose probability is greater than a probability threshold is determined as the background pixel; and the area formed by each background pixel in the depth of field blurred image is determined as the background area.

[0021] In the above scheme, the above determination module is also used to obtain an initial prediction model and an image sample set, wherein the image sample set includes multiple image samples to be identified and background area samples corresponding to each of the image samples to be identified; the following processing is performed for each of the image samples to be identified: calling the initial prediction model, performing background prediction on each pixel point sample in the image sample to be identified, and obtaining the probability that each pixel point sample is the background pixel point; determining each pixel point sample whose probability is greater than the probability threshold as the background pixel point, and determining the area formed by each of the background pixel points in the image sample to be identified as the predicted background area corresponding to the image sample to be identified; combining the predicted background area and the background area samples, training the initial prediction model to obtain the target prediction model.

[0022] In the above scheme, the above determination module is also used to determine the union of the predicted background area and the background area samples as the first reference area, and determine the intersection of the predicted background area and the background area samples as the second reference area; determine the ratio of the area of ​​the second reference area to the area of ​​the first reference area as the loss value of the initial prediction model; based on the loss value of the initial prediction model, update the model parameters of the initial prediction model to obtain the target prediction model.

[0023] In the above scheme, the above blur module is also used to perform the following processing for each pixel in the background area: obtain pixel parameters of each adjacent pixel of the pixel, and the pixel parameters include at least one of color value, transparency, and brightness value; perform weighted averaging on the values ​​of the pixel parameters of each adjacent pixel to obtain the target value of the pixel parameter; update the value of the pixel parameter of each pixel in the background area to the corresponding target value to obtain the target blurred image corresponding to the image to be identified.

[0024] In the above scheme, the above recognition module is also used to obtain a biometric information library, which includes object biometric information of multiple objects; compare the biometric information included in the foreground of the target blurred image with the biometric information of each object in the biometric information library, respectively, to obtain a comparison result corresponding to each object biometric information, the comparison result is used to indicate the degree of similarity between the object biometric information and the biometric information; based on the comparison result corresponding to each object biometric information, determine the corresponding recognition result.

[0025] In the above scheme, the above-mentioned recognition module is also used to perform the following processing respectively for the comparison results corresponding to each of the object biometric information: when the comparison result indicates that the similarity between the biometric information and the object biometric information is greater than a similarity threshold, the object corresponding to the object biometric information is determined as the target object; based on the target object, the corresponding recognition result is generated, wherein the recognition result is used to indicate that the biometric information included in the target blurred image belongs to the target object.

[0026] In the above scheme, the above acquisition module is also used to respond to a shooting instruction for an image acquisition device, acquire an image through the image acquisition device, and obtain a candidate image; perform quality detection on the candidate image to obtain a quality score of the candidate image, and the quality score is positively correlated with the quality of the biological information included in the candidate image; when the quality score is greater than a quality score threshold, the candidate image is determined as the image to be identified.

[0027] In the above scheme, the above acquisition module is also used to obtain multiple quality assessment dimensions and the weights corresponding to each quality assessment dimension; for each quality assessment dimension, the candidate image is subjected to quality detection of the quality assessment dimension to obtain the dimension quality score corresponding to the quality assessment dimension; each dimension quality score is multiplied by the corresponding weight to obtain the product result corresponding to each dimension quality score; each product result is added to obtain the quality score of the candidate image.

[0028] An embodiment of the present application provides an electronic device, including:

[0029] A memory for storing computer executable instructions or computer programs;

[0030] The processor is used to implement the biometric information recognition method provided in the embodiment of the present application when executing the computer executable instructions or computer programs stored in the memory.

[0031] The embodiment of the present application provides a computer-readable storage medium storing computer-executable instructions for causing a processor to execute the instructions to implement the biometric information recognition method provided in the embodiment of the present application.

[0032] The embodiment of the present application provides a computer program product, which includes a computer program or a computer executable instruction, and the computer program or the computer executable instruction is stored in a computer-readable storage medium. The processor of the electronic device reads the computer executable instruction from the computer-readable storage medium, and the processor executes the computer executable instruction, so that the electronic device performs the above-mentioned biometric information recognition method of the embodiment of the present application.

[0033] The embodiments of the present application have the following beneficial effects:

[0034] By acquiring an image to be identified, blurring the image to be identified through depth of field to obtain a depth-blurred image, determining the background area corresponding to the background of the depth-blurred image, blurring the background area in the depth-blurred image to obtain a target blurred image corresponding to the image to be identified, and identifying the biometric information based on the target blurred image to obtain a corresponding identification result. In this way, by blurring the image to be identified through depth of field blurring and blurring the background area, the blurring effect of the overlapping area of ​​the above two blurring processes in the image to be identified is significantly improved, thereby effectively improving the privacy security of biometric information identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 is a schematic diagram of the architecture of a system for biometric information identification provided in an embodiment of the present application;

[0036] Figure 2 is a schematic diagram of the structure of an electronic device for identifying biological information provided by an embodiment of the present application;

[0037] Figures 3 to 10 is a flow chart of a biometric information identification method provided in an embodiment of the present application;

[0038] Fig.11 It is a schematic diagram of the interactive process of the biometric information recognition method provided in the embodiment of the present application. DETAILED DESCRIPTION

[0039] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present application.

[0040] In the following description, reference is made to “some embodiments”, which describe a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0041] In the following description, the terms "first\second\third" involved are merely used to distinguish similar objects and do not represent a specific ordering of the objects. It can be understood that "first\second\third" can be interchanged with a specific order or sequence where permitted, so that the embodiments of the present application described here can be implemented in an order other than that illustrated or described here.

[0042] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0043] Before further describing the embodiments of the present application in detail, the nouns and terms involved in the embodiments of the present application are explained. The nouns and terms involved in the embodiments of the present application are subject to the following interpretations.

[0044] 1) Artificial Intelligence (AI): It is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. Artificial intelligence technology is an interdisciplinary subject that covers a wide range of fields, including both hardware-level technology and software-level technology. Basic AI technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics and other technologies.

[0045] 2) Convolutional Neural Networks (CNN): It is a type of feed-forward neural network (FNN) that includes convolution calculations and has a deep structure. It is one of the representative algorithms of deep learning. Convolutional neural networks have representation learning capabilities and can perform shift-invariant classification on input images according to their hierarchical structure.

[0046] 3) Convolutional layer: Each convolutional layer in a convolutional neural network consists of several convolutional units, and the parameters of each convolutional unit are optimized through the back propagation algorithm. The purpose of the convolution operation is to extract different features of the input. The first convolutional layer may only extract some low-level features such as edges, lines, and corners. More layers of the network can iteratively extract more complex features from low-level features.

[0047] 4) Pooling layer: After the convolution layer performs feature extraction, the output feature map is passed to the pooling layer for feature selection and information filtering. The pooling layer contains a pre-set pooling function, which replaces the result of a single point in the feature map with the feature map statistics of its adjacent area. The pooling layer selects the pooling area in the same way as the convolution kernel scans the feature map, which is controlled by the pooling size, step size, and padding.

[0048] 5) Fully-Connected Layer: The fully connected layer in a convolutional neural network is equivalent to the hidden layer in a traditional feedforward neural network. The fully connected layer is located at the end of the hidden layer of the convolutional neural network and only transmits signals to other fully connected layers. The feature map loses its spatial topological structure in the fully connected layer, is expanded into a vector and passes through the activation function.

[0049] 6) In response to: used to indicate the conditions or states on which the executed operations depend. When the dependent conditions or states are met, one or more operations executed may be in real time or have a set delay. Unless otherwise specified, there is no restriction on the order in which the multiple operations executed are executed.

[0050] 7) Bioinformation: Bioinformation is information that reflects the state and mode of biological movement. Bioinformation includes palm print information, eyeball information, fingerprint information, etc. After a long period of evolution, nature has produced organisms and gradually formed a complex biological world. Bioinformation is of various forms and ever-changing, and different types of organisms send out different information. People have made some considerable achievements in the study of bioinformation. People have also found that there is a certain inevitable connection between biological information and non-biological information. For example, the flying of swallows and wild geese indicates the change of seasons and the rise and fall of temperature; the floating of fish indicates that heavy rain is coming, and some abnormal phenomena of animals indicate that an earthquake is about to occur.

[0051] During the implementation of the embodiments of the present application, the applicant discovered that the related technology has the following problems:

[0052] In related technologies, biometric information recognition is usually performed directly on the image to be recognized. Since the image to be recognized contains a large amount of biological privacy information, if the biometric information is directly recognized on the image to be recognized, the biological privacy information will be leaked, resulting in poor privacy security of biometric information recognition.

[0053] The embodiments of the present application provide a biometric information recognition method, device, electronic device, computer-readable storage medium and computer program product, which can effectively improve the privacy security of biometric information recognition. The following describes an exemplary application of the biometric information recognition system provided by the embodiments of the present application.

[0054] See also Figure 1 , Figure 1 It is a schematic diagram of the architecture of a system 100 for biometric information identification provided in an embodiment of the present application. A terminal (terminal 400 is shown as an example) is connected to a server 200 via a network 300. The network 300 may be a wide area network or a local area network, or a combination of the two.

[0055] The terminal 400 is used for the user to use the client 410, and the recognition result is displayed on the graphical interface 410-1 (graphic interface 410-1 is shown as an example). The terminal 400 and the server 200 are connected to each other via a wired or wireless network.

[0056] In some embodiments, the server 200 may be 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, content delivery networks (CDN), and big data and artificial intelligence platforms. The terminal 400 may be a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart TV, a smart watch, a car terminal, etc., but is not limited thereto. The electronic device provided in the embodiment of the present application may be implemented as a terminal or as a server. The terminal and the server may be directly or indirectly connected via wired or wireless communication, which is not limited in the embodiment of the present application.

[0057] In some embodiments, the terminal 400 obtains an image to be identified and sends the image to be identified to the server 200. The server 200 performs precision blurring on the image to be identified to obtain a depth-of-field blurred image corresponding to the image to be identified, determines the background area corresponding to the background of the depth-of-field blurred image, and blurs the background area in the depth-of-field blurred image to obtain a target blurred image corresponding to the image to be identified. Based on the target blurred image, the biometric information is identified to obtain a corresponding identification result, and the identification result is sent to the terminal 400.

[0058] In other embodiments, the server 200 obtains an image to be identified, performs depth blur on the image to be identified, obtains a depth of field blurred image corresponding to the image to be identified, determines a background area corresponding to the background of the depth of field blurred image, and blurs the background area in the depth of field blurred image to obtain a target blurred image corresponding to the image to be identified, identifies the biometric information based on the target blurred image, obtains a corresponding identification result, and sends the identification result to the terminal 400.

[0059] In other embodiments, the embodiments of the present application can be implemented with the aid of cloud technology. Cloud technology refers to a hosting technology that unifies a series of resources such as hardware, software, and networks within a wide area network or a local area network to achieve data computing, storage, processing, and sharing.

[0060] Cloud technology is a general term for network technology, information technology, integration technology, management platform technology, and application technology based on the cloud computing business model. It can form a resource pool that can be used on demand and is flexible and convenient. Cloud computing technology will become an important support. The background services of the technical network system require a large amount of computing and storage resources.

[0061] See also Figure 2 , Figure 2 is a schematic diagram of the structure of an electronic device 500 for identifying biological information provided by an embodiment of the present application, wherein: Figure 2 The electronic device 500 shown may be Figure 1 The server 200 or the terminal 400 in Figure 2 The electronic device 500 shown includes: at least one processor 410, a memory 450, and at least one network interface 420. The various components in the electronic device 500 are coupled together via a bus system 440. It is understood that the bus system 440 is used to achieve connection and communication between these components. In addition to the data bus, the bus system 440 also includes a power bus, a control bus, and a status signal bus. However, for the sake of clarity, the bus system 440 is not described in detail. Figure 2 Various buses are labeled as bus system 440 .

[0062] Processor 410 can be an integrated circuit chip with signal processing capabilities, such as a general-purpose processor, a digital signal processor (DSP), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc., where the general-purpose processor can be a microprocessor or any conventional processor, etc.

[0063] The memory 450 may be removable, non-removable, or a combination thereof. Exemplary hardware devices include solid-state memory, hard disk drives, optical disk drives, etc. The memory 450 may optionally include one or more storage devices that are physically remote from the processor 410.

[0064] The memory 450 includes a volatile memory or a non-volatile memory, and may also include both volatile and non-volatile memories. The non-volatile memory may be a read-only memory (ROM), and the volatile memory may be a random access memory (RAM). The memory 450 described in the embodiments of the present application is intended to include any suitable type of memory.

[0065] In some embodiments, memory 450 can store data to support various operations, examples of which include programs, modules, and data structures, or a subset or superset thereof, as exemplarily described below.

[0066] Operating system 451, including system programs for processing various basic system services and performing hardware-related tasks, such as a framework layer, a core library layer, a driver layer, etc., for implementing various basic services and processing hardware-based tasks;

[0067] The network communication module 452 is used to reach other electronic devices via one or more (wired or wireless) network interfaces 420. Exemplary network interfaces 420 include: Bluetooth, Wireless Fidelity (WiFi), and Universal Serial Bus (USB).

[0068] In some embodiments, the biometric information recognition device provided in the embodiments of the present application can be implemented in software. Figure 2 The biometric information recognition device 455 stored in the memory 450 is shown, which can be software in the form of a program and a plug-in, etc., including the following software modules: an acquisition module 4551, a depth of field blur module 4552, a determination module 4553, and a blur module 4554. These modules are logical, and therefore can be arbitrarily combined or further split according to the functions implemented. The functions of each module will be described below.

[0069] In other embodiments, the biometric information identification device provided in the embodiments of the present application can be implemented in hardware. As an example, the biometric information identification device provided in the embodiments of the present application can be a processor in the form of a hardware decoding processor, which is programmed to execute the biometric information identification method provided in the embodiments of the present application. For example, the processor in the form of a hardware decoding processor can adopt one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs) or other electronic components.

[0070] In some embodiments, the terminal or server can implement the biometric information recognition method provided in the embodiment of the present application by running a computer program or a computer executable instruction. For example, the computer program can be a native program (e.g., a dedicated biometric information recognition program) or a software module in the operating system, for example, a biometric information recognition module that can be embedded in any program (such as an instant messaging client, an album program, an electronic map client, a navigation client); for example, it can be a native application (APP, Application), that is, a program that needs to be installed in the operating system to run. In short, the above-mentioned computer program can be an application, module or plug-in in any form.

[0071] The biometric information recognition method provided in the embodiment of the present application will be explained in combination with the exemplary application and implementation of the server or terminal provided in the embodiment of the present application.

[0072] See also Figure 3 , Figure 3 is a flow chart of the biometric information identification method provided in the embodiment of the present application, which will be combined with Figure 3 Steps 101 to 105 are shown for illustration. The biometric information recognition method provided in the embodiment of the present application can be implemented by a server or a terminal alone, or by a server and a terminal in collaboration. The following description will be made using the server alone as an example.

[0073] In step 101, an image to be identified is obtained, wherein the foreground of the image to be identified includes biological information.

[0074] In some embodiments, the image to be identified is any image including biological information in the foreground. Biological information is information reflecting the state and mode of biological movement, including palm print information, eyeball information, fingerprint information, etc.

[0075] In some embodiments, see Figure 4 , Figure 4 is a flow chart of a biometric information identification method provided in an embodiment of the present application, Figure 3 The illustrated step 102 may be performed by Figure 4 Steps 1011 to 1013 are implemented as shown.

[0076] In step 1011, in response to a shooting instruction for an image acquisition device, an image acquisition device is used to acquire an image to obtain a candidate image.

[0077] In some embodiments, the image acquisition device includes a camera, a camcorder, a scanner, and other devices with a camera function (mobile phones, tablet computers, smart wearable devices, etc.). The image acquisition device is an electronic device used to acquire images to be identified.

[0078] As an example, in response to a shooting instruction for an image acquisition device, an image of a biological entity or a part of a biological entity is acquired by the image acquisition device to obtain a candidate image.

[0079] As an example, in response to a shooting instruction for an image acquisition device, the image acquisition device is used to acquire an image of a palm of a living being to obtain a candidate image.

[0080] In step 1012, quality detection is performed on the candidate image to obtain a quality score of the candidate image.

[0081] In some embodiments, the quality score is positively correlated with the quality of the biological information included in the candidate image. The quality score is used to indicate the quality of the biological information included in the candidate image. The quality of the biological information can be measured by quality indicators such as the completeness and clarity of the biological information.

[0082] In some embodiments, quality detection is used to measure the quality of the biological information in the candidate image.

[0083] In some embodiments, the above step 1012 can be implemented in the following manner: obtain multiple quality assessment dimensions and the weight corresponding to each quality assessment dimension; for each quality assessment dimension, perform quality detection on the candidate image in the quality assessment dimension to obtain the dimension quality score corresponding to the quality assessment dimension; multiply the quality score of each dimension by the corresponding weight respectively to obtain the product result corresponding to the quality score of each dimension; add the product results to obtain the quality score of the candidate image.

[0084] In some embodiments, the quality assessment dimensions are used to measure the quality of the candidate images from a corresponding single dimension.

[0085] As an example, when the biometric information is palmprint information, the quality assessment dimensions include palmprint completeness, palmprint occlusion, palmprint flatness, palmprint shadow area, palmprint curvature, palmprint inclination angle, palmprint clarity and other assessment dimensions.

[0086] Continuing from the above example, the quality score of the candidate image can be expressed as:

[0087] F=α1F1+α2F2+α3F3+α4F4+α5F5+α6F6+α7F7 (1)

[0088] Among them, F is used to indicate the quality score of the candidate image, F1 is used to indicate the dimensional quality score corresponding to the palmprint completeness, F2 is used to indicate the dimensional quality score corresponding to the palmprint occlusion, F3 is used to indicate the dimensional quality score corresponding to the palmprint flatness, F4 is used to indicate the dimensional quality score corresponding to the palmprint shadow area, F5 is used to indicate the dimensional quality score corresponding to the palmprint shadow area, F6 is used to indicate the dimensional quality score corresponding to the palmprint curvature, F6 is used to indicate the dimensional quality score corresponding to the palmprint inclination angle, F7 is used to indicate the dimensional quality score corresponding to the palmprint clarity, α1, α2, α3, α4, α5, α6, and α7 are used to indicate the weights corresponding to palmprint completeness, palmprint occlusion, palmprint flatness, palmprint shadow area, palmprint curvature, palmprint inclination angle, and palmprint clarity, respectively.

[0089] In step 1013, when the quality score is greater than the quality score threshold, the candidate image is determined as the image to be identified.

[0090] In some embodiments, the quality score threshold may be specifically set according to actual application practices, and the numerical value of the quality score threshold does not constitute a limitation on the embodiments of the present application.

[0091] In this way, by performing quality detection on the candidate images captured by the image acquisition device, the images to be identified whose quality scores are higher than the quality score threshold are screened out, thereby effectively improving the image quality of the images to be identified, allowing the subsequent identification of high-quality images to be identified, thereby effectively improving the recognition accuracy of the images to be identified.

[0092] In step 102, depth-of-field blurring is performed on the image to be identified to obtain a depth-of-field blurred image corresponding to the image to be identified.

[0093] In some embodiments, depth of field blur refers to a process of blurring an image. The depth of field blur can blur the foreground and background of an image while emphasizing the clarity of the foreground.

[0094] In some embodiments, the blurriness of the depth-of-field blurred image is greater than the blurriness of the image to be recognized, and the clarity of the depth-of-field blurred image is less than the clarity of the image to be recognized.

[0095] In some embodiments, see Figure 5 , Figure 5 is a flow chart of a biometric information identification method provided in an embodiment of the present application, Figure 3 The illustrated step 102 may be performed by Figure 5 Steps 1021A to 1024A are implemented as shown.

[0096] In step 1021A, the target pixel point in the image to be identified is obtained.

[0097] In some embodiments, the target pixel point is a pixel point in the image to be identified that coincides with the focus of the image acquisition device that acquires the image to be identified.

[0098] In step 1022A, the depth of field distance is obtained.

[0099] In some embodiments, the depth of field distance is used to indicate a distance threshold between a target pixel and other pixels.

[0100] In some embodiments, the depth of field distance is less than the maximum value of the distances between the target pixel and other pixel points in the image to be identified, and the other pixel points are pixel points in the image to be identified except the target pixel point.

[0101] In step 1023A, for each other pixel point, when the distance between the other pixel point and the target pixel point is greater than the depth of field distance, the other pixel point is determined as a depth of field pixel point.

[0102] As an example, for other pixel point A and other pixel point B, the distance between other pixel point A and the target pixel point is greater than the depth of field distance, and the distance between other pixel point B and the target pixel point is less than the depth of field distance, and other pixel point A is determined as a depth of field pixel point.

[0103] In step 1024A, each depth-of-field pixel in the image to be identified is blurred to obtain a depth-of-field blurred image.

[0104] In some embodiments, the blur degree of each depth of field pixel in the image to be identified is smaller than the blur degree of the corresponding pixel in the depth of field blurred image, and the clarity of each depth of field pixel in the image to be identified is greater than the clarity of the corresponding pixel in the depth of field blurred image.

[0105] In some embodiments, see Figure 6 , Figure 6 is a flow chart of a biometric information identification method provided in an embodiment of the present application, Figure 3 The illustrated step 102 may be performed by Figure 6 Steps 1021B to 1023B are implemented as shown.

[0106] In step 1021B, the depth value of each pixel in the image to be identified and the depth threshold are obtained.

[0107] In some embodiments, the depth value of a pixel is used to indicate the image depth of the pixel, and the image depth refers to the number of bits actually required to store the grayscale or color of the image in the pixel depth. Assuming that the pixel depth of the image is 16 bits, but the number of bits used to represent the grayscale or color of the image is only 15 bits, the image depth of the image is 15. The image depth determines the number of possible colors for each pixel of the image, or the number of possible grayscale levels. For example, each pixel of a color image is represented by three components of RGB, each component is 8 bits, and the pixel depth is 24 bits. Each pixel of a digitized image is described by a set of binary numbers, which includes bits representing the color of the image. The number of binary bits occupied by the color of each pixel in the image is called the image depth. The image depth determines the number of possible colors for each pixel of a color image, or the number of possible grayscale levels for each pixel of a grayscale image. It determines the maximum number of colors that can appear in a color image, or the maximum grayscale level in a grayscale image. For a bitmap, image depth is a constant. Image depth determines the maximum number of colors that can be used in an image. In a bitmap, if each pixel has only one color bit, the pixel is either dark or bright, that is, a monochrome image (note that this is not necessarily a black and white image, it just limits the image to only two chromaticities or colors). If each pixel has 4 color bits, the bitmap supports 16 colors; if each pixel has 8 color bits, the bitmap can support 256 different colors. Image depth is used to describe the grayscale value or number of colors of each pixel in the image depth information. In the monocular image depth estimation problem, the depth information of an image corresponding to a scene is generally described by a grayscale image of the same size. The grayscale value of each pixel in the grayscale image describes the depth value of the scene corresponding to the point. The grayscale image is also called a depth map. Only by knowing the pixel depth can the depth value corresponding to each pixel be calculated.

[0108] In step 1022B, for each pixel point, when the depth value of the pixel point is greater than the depth threshold, the pixel point is determined as a depth of field pixel point.

[0109] As an example, for pixel A and pixel B, the depth threshold is 100, the depth value of pixel A is 95, and the depth value of pixel B is 102, then pixel B is determined as a depth of field pixel.

[0110] In step 1023B, each depth-of-field pixel in the image to be identified is blurred to obtain a depth-of-field blurred image.

[0111] In some embodiments, blur processing refers to the process of reducing the clarity of corresponding pixel points, the blur degree of each depth of field pixel point in the image to be identified is less than the blur degree of the corresponding pixel point in the depth of field blurred image, and the clarity of each depth of field pixel point in the image to be identified is greater than the clarity of the corresponding pixel point in the depth of field blurred image.

[0112] In this way, by performing depth blurring on the image to be identified, a depth blurred image corresponding to the image to be identified is obtained, thereby achieving preliminary blurring of the background in the image to be identified, which significantly improves the privacy security of the background in the image to be identified and can effectively enhance the privacy security of biometric information recognition.

[0113] In step 103, a background area corresponding to the background of the depth-of-field blurred image is determined.

[0114] In some embodiments, the depth-of-field blurred image includes a background area corresponding to the background and a foreground area determined by the foreground.

[0115] In some embodiments, see Figure 7 , Figure 7 is a flow chart of a biometric information identification method provided in an embodiment of the present application, Figure 3 The illustrated step 103 may be performed by Figure 7 Steps 1031 to 1034 are implemented as shown.

[0116] In step 1031, a target prediction model is obtained.

[0117] In some embodiments, the target prediction model may be a neural network model of any structure. For example, the target prediction model may be a Faster R-CNN model, a YOLO model, etc. The network structure of the target prediction model does not constitute a limitation on the embodiments of the present application.

[0118] In some embodiments, the above step 1031 can be implemented in the following manner: obtain an initial prediction model and an image sample set, the image sample set including multiple image samples to be identified and background area samples corresponding to each image sample to be identified; perform the following processing for each image sample to be identified: call the initial prediction model, perform background prediction on each pixel sample in the image sample to be identified, and obtain the probability that each pixel sample is a background pixel; determine each pixel sample whose probability is greater than the probability threshold as a background pixel, and determine the area formed by each background pixel in the image sample to be identified as the predicted background area corresponding to the image sample to be identified; train the initial prediction model in combination with the predicted background area and the background area samples to obtain the target prediction model.

[0119] In some embodiments, the numerical value of the probability threshold can be specifically set according to the actual application scenario, and the numerical value of the probability threshold does not constitute a limitation on the embodiments of the present application.

[0120] As an example, for pixel sample C and pixel sample D in the image sample to be identified, the initial prediction model is called to perform background prediction on pixel sample C in the image sample to be identified, and the probability of each pixel sample C being a background pixel is obtained; when the probability corresponding to pixel sample C is greater than the probability threshold, pixel sample C is determined to be a background pixel. The initial prediction model is called to perform background prediction on pixel sample D in the image sample to be identified, and the probability of pixel sample D being a background pixel is obtained; when the probability corresponding to pixel sample D is greater than the probability threshold, pixel sample D is determined to be a background pixel.

[0121] In some embodiments, the above-mentioned combination of predicted background area and background area samples is used to train the initial prediction model to obtain a target prediction model, which can be achieved in the following way: the union of the predicted background area and the background area samples is determined as the first reference area, and the intersection of the predicted background area and the background area samples is determined as the second reference area; the ratio of the area of ​​the second reference area to the area of ​​the first reference area is determined as the loss value of the initial prediction model; based on the loss value of the initial prediction model, the model parameters of the initial prediction model are updated to obtain the target prediction model.

[0122] As an example, the loss value of the above initial prediction model can be expressed as:

[0123]

[0124] Wherein, L is used to indicate the loss value of the initial prediction model, S1 is used to indicate the area of ​​the second reference area, and S2 is used to indicate the area of ​​the first reference area.

[0125] In other embodiments, the above step 1031 can be implemented in the following manner: obtaining an initial prediction model and an image sample set, the image sample set including multiple image samples to be identified and sample labels of each image sample to be identified, the sample label being used to indicate whether each pixel in the image to be identified is a pixel in the background area; performing the following processing for each image sample to be identified: calling the initial prediction model, performing background prediction on each pixel sample in the image sample to be identified, and obtaining a prediction result for each pixel sample, the prediction result being used to indicate whether the pixel sample is a pixel in the background area; training the initial prediction model in combination with the prediction result and the sample label to obtain a target prediction model.

[0126] In this way, by training the initial prediction model, the obtained target prediction model can achieve accurate prediction of pixel points, effectively improving the model performance of the target prediction model, thereby effectively improving the accuracy of background prediction.

[0127] In step 1032, the target prediction model is called to perform background prediction on each pixel in the depth of field blurred image to obtain the probability that each pixel is a background pixel.

[0128] In some embodiments, background pixels are pixels in the background of the depth-of-field blurred image.

[0129] In step 1033, each pixel point having a probability greater than a probability threshold is determined as a background pixel point.

[0130] As an example, for pixel E and pixel F, the probability corresponding to pixel E is 0.8, the probability corresponding to pixel F is 0.4, and the probability threshold is 0.6, so pixel E can be determined as a background pixel.

[0131] In step 1034, the area formed by the background pixels in the depth of field blurred image is determined as the background area.

[0132] In some embodiments, each background pixel in the depth of field blurred image includes background pixel E, background pixel F and background pixel G, and the area formed by background pixel E, background pixel F and background pixel G in the depth of field blurred image is determined as the background area.

[0133] In this way, by calling the target prediction model, the background prediction is performed on each pixel in the depth of field blurred image to determine the background pixel points in the depth of field blurred image, and the area formed by the background pixel points is determined as the background area. Since the target prediction model can accurately predict the background pixel points, the accuracy of the determined background area is higher, thereby effectively improving the accuracy of the determined background area.

[0134] In step 104, the background area in the depth-of-field blurred image is blurred to obtain a target blurred image corresponding to the image to be recognized.

[0135] In some embodiments, the blur degree of the background area in the depth of field blurred image is less than the blur degree of the target blurred image corresponding to the image to be identified, and the blur processing is used to reduce the clarity of the background area in the depth of field blurred image.

[0136] In some embodiments, see Figure 8 , Figure 8 is a flow chart of a biometric information identification method provided in an embodiment of the present application, Figure 3 Step 104 shown may be performed by Figure 8 Steps 1041 to 1043 are implemented as shown.

[0137] In step 1041, the following processing is performed for each pixel in the background area: pixel parameters of each adjacent pixel of the pixel are obtained.

[0138] In some embodiments, the pixel parameter includes at least one of a color value, a transparency, and a brightness value.

[0139] In some embodiments, the color value is used to indicate the color of the corresponding pixel, the transparency is used to indicate the degree of transparency of the corresponding pixel, and the brightness value is used to indicate the brightness of the corresponding pixel.

[0140] In step 1042, the values ​​of the pixel parameters of the adjacent pixels are weighted averaged to obtain the target value of the pixel parameter.

[0141] As an example, the expression for the target value of a pixel parameter may be:

[0142] M=β1M1+β2M2+…+β n M n (3)

[0143] Among them, M is used to indicate the target value of the pixel parameter, M1, M2…M n It is used to indicate the pixel parameters of adjacent pixels, and n is used to indicate the number of adjacent pixels.

[0144] In step 1043, the value of the pixel parameter of each pixel point in the background area is updated to the corresponding target value, so as to obtain a target blurred image corresponding to the image to be recognized.

[0145] In this way, by taking a weighted average of the values ​​of the pixel parameters of each adjacent pixel of the pixel point, a target value of the pixel parameter is obtained, and the value of the pixel parameter of the pixel point is updated to the corresponding target value, thereby achieving blurring of the background area in the depth of field blurred image, thereby achieving secondary blurring of the background in the image to be identified, further improving the privacy security of the background in the image to be identified, and effectively improving the accuracy of the determined background area.

[0146] In step 105, the biological information is recognized based on the target blurred image to obtain a corresponding recognition result.

[0147] In some embodiments, the recognition result is used to indicate whether the biological information included in the target blurred image belongs to the target object.

[0148] In some embodiments, see Fig. 9 , Fig. 9 is a flow chart of a biometric information identification method provided in an embodiment of the present application, Figure 3 The illustrated step 105 may be performed by Fig. 9 Steps 1051 to 1053 are implemented as shown.

[0149] In step 1051, a biological information database is obtained, wherein the biological information database includes subject biological information of a plurality of subjects.

[0150] In some embodiments, the subject biometric information of different objects is different, and the subject biometric information of the objects is used to identify the corresponding objects.

[0151] As an example, the biometric information database includes palm print information of object A, palm print information of object B, palm print information of object C, palm print information of object D, palm print information of object E, palm print information of object F, and the like.

[0152] In some embodiments, the biometric information library includes index entries of multiple pieces of object biometric information of objects, and the index entries are used to record the mapping relationship between the objects and the corresponding object biometric information.

[0153] In step 1052, the biological information included in the foreground of the target blurred image is compared with the biological information of each object in the biological information library to obtain the comparison results corresponding to the biological information of each object.

[0154] In some embodiments, the comparison result is used to indicate the degree of similarity between the subject biological information and the biological information.

[0155] As an example, the biometric information library includes object biometric information A, object biometric information B, object biometric information C, object biometric information D, object biometric information E, object biometric information F, and object biometric information E. The biometric information included in the foreground of the target blurred image is compared with the object biometric information A, object biometric information B, object biometric information C, object biometric information D, object biometric information E, object biometric information F, and object biometric information E, respectively, to obtain comparison results corresponding to the object biometric information A, object biometric information B, object biometric information C, object biometric information D, object biometric information E, object biometric information F, and object biometric information E, respectively.

[0156] In step 1053, a corresponding recognition result is determined based on the comparison result corresponding to each object's biological information.

[0157] In some embodiments, the above step 1053 can be implemented as follows: perform the following processing for the comparison results corresponding to the biological information of each object: when the comparison result indicates that the similarity between the biological information and the object biological information is greater than the similarity threshold, determine the object corresponding to the object biological information as the target object; based on the target object, generate a corresponding recognition result.

[0158] In some embodiments, the recognition result is used to indicate that the biological information included in the target blurred image belongs to the target object.

[0159] In this way, since the depth blurred image obtained by the first depth blur has formed a preliminary foreground and background pattern, the background area can be identified through the target prediction model, which can effectively save the computational cost of the target prediction model, thereby effectively improving the efficiency of privacy protection in the biometric information recognition process.

[0160] In this way, by acquiring an image to be identified, blurring the image to be identified by depth of field, obtaining a depth-blurred image, determining the background area corresponding to the background of the depth-blurred image, blurring the background area in the depth-blurred image, obtaining a target blurred image corresponding to the image to be identified, and based on the target blurred image, identifying the biometric information to obtain a corresponding identification result. In this way, by blurring the image to be identified by depth of field and blurring the background area, the blurring effect of the overlapping area of ​​the above two blurring processes in the image to be identified is significantly improved, effectively improving the privacy security of biometric information identification.

[0161] Next, an exemplary application of the embodiment of the present application in an actual palmprint information recognition application scenario will be described.

[0162] Palm prints refer to the image of the palm from the end of the finger to the wrist. Many of these features can be used for identity recognition: such as main lines, wrinkles, fine textures, ridge ends, bifurcation points, etc. Palm print recognition is also a non-invasive recognition method that is easy for users to accept and does not require high collection equipment. Palm print recognition can be applied to the field of mobile payment. By collecting biological palm prints, palm veins and other characteristic information, it can realize offline payment, access control, identification, security check and other functions. For example, users can use face recognition devices to scan their palms. They need to bind their personal account on the device and enter the palm print. When consuming, they only need to align their palms with the scanning area of ​​the payment device to confirm and complete the payment.

[0163] Palmprint recognition, as a biometric technology, has the advantages of being safe, fast, and convenient, and is widely used in payment, access control, service and other scenarios. However, palmprint recognition also has the risk of privacy leakage, especially the problem of background information leakage. The embodiment of the present application uses a camera to shoot a palmprint image, blurs the background information of the palmprint image through depth of field blur, and uses a convolutional neural network to perform a secondary blur on the image, thereby achieving effective privacy protection of the palmprint background information. The embodiment of the present application applies depth of field blur and convolutional neural network to the field of palmprint recognition, and protects the privacy security of users by improving the privacy protection capability of the palmprint background. Below, the depth of field blur and the blurring method using a convolutional neural network will be described respectively.

[0164] In some embodiments, when taking a palm image, the following steps can be used to achieve depth of field blur: Adjust the focal length of the shooting device to the maximum: Adjust the focal length of the camera or mobile phone to the maximum, so that the depth of field range of the camera can be minimized, making the palm print image clear and the background blurred. Adjust the aperture of the camera to the maximum: Adjust the aperture of the camera to the maximum to maximize the amount of light entering the camera, making the palm print image clearer. Adjust the distance of the palm: Adjust the distance between the palm and the camera to the minimum, so that the palm print image can be clearer. Adjust the position of the camera: When taking a palm print image, point the camera at the palm so that the camera can fully capture the palm print image. Post-processing: After taking the palm print image, you can use post-processing tools to perform depth of field blur processing to protect the user's privacy. By blurring the background of the palm print image, the risk of background information leakage can be reduced while maintaining the clarity of the palm, thereby protecting the user's privacy. This patent uses image background blurring technology based on convolutional neural networks.

[0165] In some embodiments, for blurring using a convolutional neural network, when a palm image is captured using a camera, a depth of field blur is first performed, and then a convolutional neural network is used to locate the background information in the palm image. Next, the background information is blurred by using a blur filter in the background area, making the background difficult to identify, thereby ensuring the privacy and security of the user. The following are the specific steps: First, depth of field blurring: When a palm image is captured using a camera, depth of field blurring can be achieved by controlling the focal length of the lens, that is, blurring the background, thereby reducing the risk of background information leakage. This step is usually completed by the camera hardware and does not require additional algorithm support. Use a convolutional neural network to locate the background information in the palm image: In the captured image, a convolutional neural network is used to process the image and extract the background information in the image. Specifically, a target detection algorithm based on deep learning, such as Faster R-CNN, YOLO, etc., can be used to identify the area where the background information is located. Then, the background information is blurred: After the background information is located, a blur filter is used to blur it, making it difficult to identify the background information. Here, Gaussian filters, mean filters, etc. can be used, and appropriate filters are selected according to actual conditions. Through the above steps, the embodiment of the present application can effectively protect the user's privacy information while ensuring the quality and clarity of the palm print image.

[0166] In some embodiments, see Fig.10 , Fig.10 The flowchart of the biological information recognition method provided in the embodiment of the present application is shown in FIG. The training of the convolutional neural network can be performed by Fig.10 Steps 201 to 216 are shown to be implemented.

[0167] In step 201, a palmprint image dataset is collected.

[0168] In step 202, the data set is divided and preprocessed.

[0169] In step 203, a convolutional neural network structure is constructed.

[0170] In some embodiments, a suitable convolutional neural network structure is selected, such as a deep residual network (ResNet) or a lightweight network (MobileNet), etc. The network structure is adjusted and optimized according to application requirements.

[0171] In step 204, it is determined whether the image in the palmprint image data set is overexposed or of unqualified quality. If so, step 205 is executed; if not, step 206 is executed.

[0172] In step 205, data screening is performed.

[0173] In step 206, training data is prepared.

[0174] In some embodiments, the collected palmprint image data set and corresponding label information are processed and prepared, such as data set division, data enhancement, etc. At the same time, the background of the palmprint image is processed so that the background of the palmprint image in the training set and the test set has a consistent blur level.

[0175] In step 207, the network structure is adjusted.

[0176] In step 208, the model is retrained.

[0177] In step 209, network parameters are initialized.

[0178] In some embodiments, various parameters of the convolutional neural network, such as weights and biases, are initialized.

[0179] In step 210, forward propagation and loss calculation.

[0180] In some embodiments, the input palm print image is sent to the convolutional neural network through forward propagation for feature extraction and classification. The loss value is calculated through the loss function for subsequent back propagation and gradient update.

[0181] In step 211, back propagation and gradient update.

[0182] In some embodiments, a back-propagation algorithm is used to calculate the gradients of various parameters of the network, and the parameters are updated according to the gradients to make the loss function as small as possible.

[0183] In step 212, it is determined whether the model performance is excellent. If it is excellent, step 213 is executed; otherwise, step 214 is executed.

[0184] In step 213, a model test is performed.

[0185] In step 214, hyperparameters are adjusted.

[0186] In step 215, a regularization technique is used.

[0187] In some embodiments, the validation set is used to evaluate and optimize the trained model, such as adjusting hyperparameters, using regularization techniques, etc. At the same time, for the background of the palm print image, the parameters of the model are optimized to further improve the blurriness of the background and protect user privacy.

[0188] In step 216, the model is retrained.

[0189] In some embodiments, for targeted tuning:

[0190] For palm images with excessive exposure or unqualified quality, the embodiments of the present application can perform fine-tuning to optimize the performance of the model. The specific steps may include the following aspects: Collect palmprint image data sets: For palm images with excessive exposure or unqualified quality, the embodiments of the present application can collect some representative data sets in a targeted manner, such as simulated overexposed images, simulated blurred images, etc. Data processing and preparation: Preprocess and prepare the collected data, such as image enhancement, data set division, labeling, etc. Transfer learning and fine-tuning: Apply the pre-trained convolutional neural network model to the new data set, perform transfer learning and fine-tune to improve the performance of the model on the new data set. Model evaluation and optimization: Use the validation set to evaluate and optimize the fine-tuned model, such as adjusting hyperparameters, using regularization techniques, etc. Through the above fine-tuning steps, the embodiments of the present application can optimize the performance of the convolutional neural network, making its recognition ability for palm images with excessive exposure or unqualified quality more accurate and stable.

[0191] In some embodiments, for the biometric information recognition method provided in the embodiments of the present application, first, data collection and preprocessing: first, it is necessary to collect a large number of image data sets containing palm prints and different types of backgrounds, and perform preprocessing, such as image cropping, size standardization, and color space conversion. Then, blurring: using depth of field blurring technology to blur the background, the background depth of field information can be obtained through a depth sensor or an ordinary camera, and then blurring is performed according to the algorithm, and finally an image with only palm prints is generated. Then, convolutional neural network training: using deep learning technology, the palm print image is trained through the convolutional neural network to generate an efficient palm print recognition model. The training data includes palm print images and corresponding label information. Finally, model application: deploy the trained convolutional neural network model on the mobile device to identify the user's palm prints. During the recognition process, the background is firstly blurred with depth of field to protect the user's privacy, and then the processed palm print image is input into the convolutional neural network for recognition and verification.

[0192] In some embodiments, see Fig.11 , Fig.11 The interactive flow diagram of the biometric information recognition method provided in the embodiment of the present application is as follows. Fig.11 Steps 301 to 313 shown implement image acquisition and registration.

[0193] In step 301, check whether the camera permission is enabled, and if enabled, execute step 302; otherwise, execute step 303.

[0194] In step 302, the camera is turned on.

[0195] In step 303, the user is prompted to enable camera permission.

[0196] In step 304, it is determined whether the focusing is successful. If the focusing is successful, step 305 is executed; if the focusing is unsuccessful, step 306 is executed.

[0197] In step 305, a palm photo is taken.

[0198] In step 306, the user is prompted that the focusing has failed.

[0199] In step 307, it is determined that the palm quality is qualified, and if qualified, step 308 is executed; otherwise, step 309 is executed.

[0200] In step 308, depth of field blur processing is performed.

[0201] In step 309, the user is prompted to take a photo of the palm again.

[0202] In step 310, the blurred palm image is saved.

[0203] In step 311, it is determined whether the palm image is saved successfully. If successful, step 312 is executed; otherwise, step 313 is executed.

[0204] In step 312, the user is prompted that the registration is successful.

[0205] In step 313, the user is prompted that the image saving fails.

[0206] In this way, the embodiment of the present application uses the depth of field blur technology to blur the palm print background, thereby reducing the risk of background information leakage, while maintaining the clarity of the palm, thereby protecting the privacy and security of the user. The security of palm print identity recognition by users can be improved. The convolutional neural network model is introduced to perform secondary blurring on the palm print image to further protect the privacy of the user, thereby improving the accuracy and effect of the system's palm print image privacy protection.

[0207] In this way, the biometric information recognition method provided by the embodiment of the present application can protect the privacy and security of users, effectively reduce the risk of background information leakage, and improve the accuracy and efficiency of palmprint recognition. The system combines depth of field blur technology and convolutional neural network, which can automatically detect and blur the background, retaining only the palmprint image, thereby effectively protecting the privacy of users. At the same time, the system has the ability to quickly recognize palmprints without the need for additional user operations, thereby improving the user experience. In addition, the system can also adapt to the palmprint recognition needs in different environments, providing users with more comprehensive and reliable palmprint recognition and privacy protection services.

[0208] In this way, by acquiring an image to be identified, blurring the image to be identified by depth of field, obtaining a depth-blurred image, determining the background area corresponding to the background of the depth-blurred image, blurring the background area in the depth-blurred image, obtaining a target blurred image corresponding to the image to be identified, and based on the target blurred image, identifying the biometric information to obtain a corresponding identification result. In this way, by blurring the image to be identified by depth of field and blurring the background area, the blurring effect of the overlapping area of ​​the above two blurring processes in the image to be identified is significantly improved, effectively improving the privacy security of biometric information identification.

[0209] It is understandable that in the embodiments of the present application, related data such as images to be identified are involved. When the embodiments of the present application are applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of relevant data need to comply with relevant laws, regulations and standards of relevant countries and regions.

[0210] The following is a description of an exemplary structure of the biometric information recognition device 455 provided in the embodiment of the present application as a software module. In some embodiments, Figure 2As shown, the software modules stored in the biometric information recognition device 455 of the memory 450 may include: an acquisition module 4551, used to acquire an image to be recognized, wherein the foreground of the image to be recognized includes biometric information; a depth-of-field blur module 4552, used to perform depth-of-field blur on the image to be recognized, and obtain a depth-of-field blurred image corresponding to the image to be recognized; a determination module 4553, used to determine a background area corresponding to the background of the depth-of-field blurred image; a blur module 4554, used to blur the background area in the depth-of-field blurred image, and obtain a target blurred image corresponding to the image to be recognized; and an identification module 4555, used to identify the biometric information based on the target blurred image, and obtain a corresponding identification result.

[0211] In some embodiments, the above-mentioned depth of field blur module is also used to obtain a target pixel point in the image to be identified, the target pixel point is a pixel point in the image to be identified that coincides with the focus of the image acquisition device that acquires the image to be identified; obtain a depth of field distance, the depth of field distance is less than the maximum value of the distance between the target pixel point and each other pixel point in the image to be identified, the other pixel points are the pixel points in the image to be identified other than the target pixel point; for each of the other pixel points, when the distance between the other pixel point and the target pixel point is greater than the depth of field distance, the other pixel point is determined as a depth of field pixel point; and blur each of the depth of field pixel points in the image to be identified to obtain the depth of field blurred image.

[0212] In some embodiments, the above-mentioned blur module is also used to obtain the depth value of each pixel point in the image to be identified, and the depth threshold; for each pixel point, when the depth value of the pixel point is greater than the depth threshold, the pixel point is determined as a depth of field pixel point; and each depth of field pixel point in the image to be identified is blurred separately to obtain the depth of field blurred image.

[0213] In some embodiments, the above-mentioned determination module is also used to obtain a target prediction model, and call the target prediction model to perform background prediction on each pixel point in the depth of field blurred image to obtain the probability that each pixel point is a background pixel point, and the background pixel point is a pixel point in the background of the depth of field blurred image; each pixel point whose probability is greater than a probability threshold is determined as the background pixel point; and the area formed by each background pixel point in the depth of field blurred image is determined as the background area.

[0214] In some embodiments, the above-mentioned determination module is also used to obtain an initial prediction model and an image sample set, wherein the image sample set includes multiple image samples to be identified and background area samples corresponding to each of the image samples to be identified; the following processing is performed for each of the image samples to be identified: calling the initial prediction model, performing background prediction on each pixel point sample in the image sample to be identified, and obtaining the probability that each pixel point sample is the background pixel point; determining each pixel point sample whose probability is greater than the probability threshold as the background pixel point, and determining the area formed by each of the background pixel points in the image sample to be identified as the predicted background area corresponding to the image sample to be identified; combining the predicted background area and the background area samples, training the initial prediction model to obtain the target prediction model.

[0215] In some embodiments, the above-mentioned determination module is also used to determine the union of the predicted background area and the background area samples as the first reference area, and determine the intersection of the predicted background area and the background area samples as the second reference area; determine the ratio of the area of ​​the second reference area to the area of ​​the first reference area as the loss value of the initial prediction model; based on the loss value of the initial prediction model, update the model parameters of the initial prediction model to obtain the target prediction model.

[0216] In some embodiments, the above-mentioned blur module is also used to perform the following processing on each pixel point in the background area: obtain pixel parameters of each adjacent pixel point of the pixel point, and the pixel parameters include at least one of color value, transparency, and brightness value; perform weighted averaging on the values ​​of the pixel parameters of each of the adjacent pixel points to obtain the target value of the pixel parameter; update the value of the pixel parameter of each of the pixel points in the background area to the corresponding target value to obtain the target blurred image corresponding to the image to be identified.

[0217] In some embodiments, the above-mentioned recognition module is also used to obtain a biometric information library, which includes object biometric information of multiple objects; compare the biometric information included in the foreground of the target blurred image with the biometric information of each object in the biometric information library, respectively, to obtain a comparison result corresponding to each object biometric information, and the comparison result is used to indicate the degree of similarity between the object biometric information and the biometric information; based on the comparison result corresponding to each object biometric information, determine the corresponding recognition result.

[0218] In some embodiments, the above-mentioned recognition module is also used to perform the following processing respectively for the comparison results corresponding to the biological information of each object: when the comparison result indicates that the similarity between the biological information and the object biological information is greater than a similarity threshold, the object corresponding to the object biological information is determined as the target object; based on the target object, the corresponding recognition result is generated, wherein the recognition result is used to indicate that the biological information included in the target blurred image belongs to the target object.

[0219] In some embodiments, the above-mentioned acquisition module is also used to respond to a shooting instruction for an image acquisition device, acquire an image through the image acquisition device, and obtain a candidate image; perform quality detection on the candidate image to obtain a quality score of the candidate image, and the quality score is positively correlated with the quality of the biological information included in the candidate image; when the quality score is greater than a quality score threshold, the candidate image is determined as the image to be identified.

[0220] In some embodiments, the above-mentioned acquisition module is also used to obtain multiple quality assessment dimensions and the weights corresponding to each quality assessment dimension; for each quality assessment dimension, the candidate image is subjected to quality detection of the quality assessment dimension to obtain the dimension quality score corresponding to the quality assessment dimension; each dimension quality score is multiplied by the corresponding weight to obtain the product result corresponding to each dimension quality score; each product result is added to obtain the quality score of the candidate image.

[0221] The embodiment of the present application provides a computer program product, which includes a computer program or a computer executable instruction, and the computer program or the computer executable instruction is stored in a computer-readable storage medium. The processor of the electronic device reads the computer executable instruction from the computer-readable storage medium, and the processor executes the computer executable instruction, so that the electronic device performs the above-mentioned biometric information recognition method of the embodiment of the present application.

[0222] The present application embodiment provides a computer-readable storage medium storing computer-executable instructions, wherein the computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, the processor will be caused to execute the biometric information recognition method provided by the present application embodiment, for example, Figure 3 The biometric information recognition method is shown.

[0223] In some embodiments, the computer-readable storage medium may be a memory such as FRAM, ROM, PROM, EPROM, EEP ROM, flash memory, magnetic surface storage, optical disk, or CD-ROM; or it may be various electronic devices including one or any combination of the above memories.

[0224] In some embodiments, computer executable instructions may be in the form of a program, software, software module, script or code, written in any form of programming language (including compiled or interpreted languages, or declarative or procedural languages), and may be deployed in any form, including as a stand-alone program or as a module, component, subroutine or other unit suitable for use in a computing environment.

[0225] As an example, computer-executable instructions may, but need not, correspond to a file in a file system, may be stored as part of a file storing other programs or data, such as in one or more scripts in a HyperText Markup Language (HTML) document, in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files storing one or more modules, subroutines, or code portions).

[0226] As an example, computer executable instructions may be deployed to be executed on one electronic device, or on multiple electronic devices located at one site, or on multiple electronic devices distributed at multiple sites and interconnected by a communication network.

[0227] In summary, the embodiments of the present application have the following beneficial effects:

[0228] (1) Obtaining an image to be identified, blurring the image to be identified by depth of field to obtain a depth-blurred image, determining the background area corresponding to the background of the depth-blurred image, blurring the background area in the depth-blurred image to obtain a target blurred image corresponding to the image to be identified, and identifying the biometric information based on the target blurred image to obtain a corresponding identification result. In this way, by blurring the image to be identified by depth of field and blurring the background area, the blurring effect of the overlapping area of ​​the above two blurring processes in the image to be identified is significantly improved, thereby effectively improving the privacy security of biometric information identification.

[0229] (2) By performing quality detection on the candidate images captured by the image acquisition device, the images to be identified whose quality scores are higher than the quality score threshold are screened out, thereby effectively improving the image quality of the images to be identified, so that the high-quality images to be identified can be subsequently identified, thereby effectively improving the recognition accuracy of the images to be identified.

[0230] (3) By performing depth blurring on the image to be identified, a depth blurred image corresponding to the image to be identified is obtained, thereby achieving preliminary blurring of the background in the image to be identified, significantly improving the privacy security of the background in the image to be identified, and effectively improving the privacy security of biometric information identification.

[0231] (4) By training the initial prediction model, the target prediction model obtained can achieve accurate prediction of pixel points, effectively improving the model performance of the target prediction model, thereby effectively improving the accuracy of background prediction.

[0232] (5) By calling the target prediction model, background prediction is performed on each pixel in the depth of field blurred image to determine the background pixels in the depth of field blurred image, and the area formed by the background pixels is determined as the background area. Since the target prediction model can achieve accurate prediction of the background pixels, the accuracy of the determined background area is high, which effectively improves the accuracy of the determined background area.

[0233] (6) By performing weighted averaging on the values ​​of pixel parameters of adjacent pixels of a pixel point, a target value of the pixel parameter is obtained, and the value of the pixel parameter of the pixel point is updated to the corresponding target value, thereby blurring the background area in the depth of field blurred image, thereby achieving secondary blurring of the background in the image to be identified, further improving the privacy security of the background in the image to be identified, and effectively improving the accuracy of the determined background area.

[0234] (7) The embodiment of the present application uses depth of field blur technology to blur the palm print background, thereby reducing the risk of background information leakage, while maintaining the clarity of the palm, thereby protecting the privacy of the user. The security of palm print identity recognition can be improved. The convolutional neural network model is introduced to perform secondary blurring on the palm print image to further protect the user's privacy, thereby improving the accuracy and effect of the system's palm print image privacy protection.

[0235] (8) The embodiment of the present application uses depth of field blur technology to blur the palm print background, thereby reducing the risk of background information leakage, while maintaining the clarity of the palm, thereby protecting the privacy of the user. The security of palm print identity recognition can be improved. The convolutional neural network model is introduced to perform secondary blurring on the palm print image to further protect the user's privacy, thereby improving the accuracy and effect of the system's palm print image privacy protection.

[0236] (9) The biometric information recognition method provided by the embodiment of the present application can protect the privacy and security of users, effectively reduce the risk of background information leakage, and improve the accuracy and efficiency of palmprint recognition. The system combines depth of field blur technology and convolutional neural network, which can automatically detect and blur the background and only retain the palmprint image, thereby effectively protecting the privacy of users. At the same time, the system has the ability to quickly recognize palmprints without the need for additional user operations, which improves the user experience. In addition, the system can also adapt to the palmprint recognition needs in different environments, providing users with more comprehensive and reliable palmprint recognition and privacy protection services.

[0237] (10) Since the depth-of-field blurred image obtained by the first depth-of-field blur has formed a preliminary foreground and background pattern, the background area can be identified through the target prediction model, which can effectively save the computational cost of the target prediction model, thereby effectively improving the efficiency of privacy protection in the biometric information recognition process.

[0238] The above is only an embodiment of the present application and is not intended to limit the protection scope of the present application. Any modifications, equivalent substitutions and improvements made within the spirit and scope of the present application are included in the protection scope of the present application.

Claims

1. A method for identifying biological information, characterized in that: The method comprises: Acquire an image to be identified, wherein the foreground of the image to be identified includes biological information; Performing depth-of-field blurring on the image to be identified to obtain a depth-of-field blurred image corresponding to the image to be identified; Determining a background area corresponding to the background of the depth-of-field blurred image; Performing blur processing on the background area in the depth-of-field blurred image to obtain a target blurred image corresponding to the image to be identified; Based on the target blurred image, the biological information is recognized to obtain a corresponding recognition result.

2. The method according to claim 1, characterized in that The step of performing depth-of-field blurring on the image to be identified to obtain a depth-of-field blurred image corresponding to the image to be identified includes: Acquire a target pixel point in the image to be recognized, wherein the target pixel point is a pixel point in the image to be recognized that coincides with a focus of an image acquisition device that acquires the image to be recognized; Acquire a depth of field distance, where the depth of field distance is less than a maximum value of distances between the target pixel and other pixel points in the image to be identified, where the other pixel points are pixel points in the image to be identified other than the target pixel point; For each of the other pixel points, when the distance between the other pixel point and the target pixel point is greater than the depth of field distance, the other pixel point is determined as a depth of field pixel point; Each of the depth-of-field pixel points in the image to be identified is blurred to obtain the depth-of-field blurred image.

3. The method according to claim 1, characterized in that: The step of performing depth-of-field blurring on the image to be identified to obtain a depth-of-field blurred image corresponding to the image to be identified includes: Obtaining a depth value of each pixel in the image to be identified and a depth threshold; For each pixel point, when the depth value of the pixel point is greater than the depth threshold, the pixel point is determined as a depth of field pixel point; Each of the depth-of-field pixel points in the image to be identified is blurred to obtain the depth-of-field blurred image.

4. The method according to claim 1, characterized in that: The determining of the background area corresponding to the background of the depth-of-field blurred image includes: Obtaining a target prediction model, calling the target prediction model, performing background prediction on each pixel in the depth-of-field blurred image, and obtaining a probability that each pixel is a background pixel, where the background pixel is a pixel in the background of the depth-of-field blurred image; Determine each pixel point whose probability is greater than a probability threshold as the background pixel point; The area formed by the background pixels in the depth-of-field blurred image is determined as the background area.

5. The method according to claim 4, characterized in that The obtaining of the target prediction model comprises: Acquire an initial prediction model and an image sample set, wherein the image sample set includes a plurality of image samples to be identified and background area samples corresponding to each of the image samples to be identified; The following processing is performed for each of the image samples to be identified: Calling the initial prediction model to perform background prediction on each pixel sample in the image sample to be identified, and obtaining the probability that each pixel sample is the background pixel; Determine each of the pixel samples whose probability is greater than the probability threshold as the background pixel, and determine the area formed by each of the background pixel points in the image sample to be identified as the predicted background area corresponding to the image sample to be identified; The initial prediction model is trained in combination with the predicted background area and the background area samples to obtain the target prediction model.

6. The method according to claim 5, characterized in that The combining the predicted background area and the background area samples to train the initial prediction model to obtain the target prediction model includes: Determine the union of the predicted background area and the background area samples as a first reference area, and determine the intersection of the predicted background area and the background area samples as a second reference area; Determine a ratio of an area of ​​the second reference area to an area of ​​the first reference area as a loss value of the initial prediction model; Based on the loss value of the initial prediction model, the model parameters of the initial prediction model are updated to obtain the target prediction model.

7. The method according to claim 1, characterized in that The blurring of the background area in the depth-of-field blurred image to obtain a target blurred image corresponding to the image to be identified includes: The following processing is performed for each pixel in the background area: obtaining pixel parameters of each adjacent pixel of the pixel, wherein the pixel parameters include at least one of a color value, a transparency, and a brightness value; Taking a weighted average of the values ​​of the pixel parameters of the adjacent pixels to obtain a target value of the pixel parameter; The value of the pixel parameter of each pixel point in the background area is updated to the corresponding target value to obtain a target blurred image corresponding to the image to be recognized.

8. The method according to claim 1, characterized in that The step of identifying the biological information based on the target blurred image to obtain a corresponding identification result includes: Acquire a biological information library, wherein the biological information library includes subject biological information of a plurality of subjects; Comparing the biological information included in the foreground of the target blurred image with the biological information of each object in the biological information library, respectively, to obtain a comparison result corresponding to each object biological information, wherein the comparison result is used to indicate the degree of similarity between the biological information of the object and the biological information; Based on the comparison results corresponding to the biological information of each object, the corresponding recognition result is determined.

9. The method according to claim 8, characterized in that The determining the corresponding recognition result based on the comparison result corresponding to each of the object biometric information includes: The following processing is performed for each comparison result corresponding to the object biological information: When the comparison result indicates that the similarity between the biological information and the object biological information is greater than a similarity threshold, determining the object corresponding to the object biological information as a target object; Based on the target object, the corresponding recognition result is generated, wherein the recognition result is used to indicate that the biological information included in the target blurred image belongs to the target object.

10. The method according to claim 1, characterized in that The step of obtaining the image to be identified comprises: In response to a shooting instruction for an image acquisition device, performing image acquisition by the image acquisition device to obtain a candidate image; Performing quality detection on the candidate image to obtain a quality score of the candidate image, wherein the quality score is positively correlated with the quality of the biological information included in the candidate image; When the quality score is greater than the quality score threshold, the candidate image is determined as the image to be identified.

11. The method according to claim 10, characterized in that The performing quality detection on the candidate image to obtain the quality score of the candidate image includes: Obtaining multiple quality assessment dimensions and weights corresponding to each of the quality assessment dimensions; For each of the quality assessment dimensions, performing quality detection of the quality assessment dimension on the candidate image to obtain a dimension quality score corresponding to the quality assessment dimension; Multiplying the quality scores of each dimension by the corresponding weight respectively to obtain a product result corresponding to the quality scores of each dimension; The multiplication results are added together to obtain the quality score of the candidate image.

12. A biometric information recognition device, characterized in that: The device comprises: An acquisition module, used for acquiring an image to be identified, wherein the foreground of the image to be identified includes biological information; A depth-of-field blur module, used to perform depth-of-field blurring on the image to be identified to obtain a depth-of-field blurred image corresponding to the image to be identified; A determination module, used to determine a background area corresponding to the background of the depth of field blurred image; A blur module, used for blurring the background area in the depth-of-field blurred image to obtain a target blurred image corresponding to the image to be identified; The recognition module is used to recognize the biological information based on the target blurred image to obtain a corresponding recognition result.

13. An electronic device, characterized in that: The electronic device comprises: A memory for storing computer executable instructions or computer programs; The processor is used to implement the biometric information recognition method according to any one of claims 1 to 11 when executing the computer executable instructions or computer programs stored in the memory.

14. A computer-readable storage medium storing computer-executable instructions, characterized in that: When the computer executable instructions are executed by a processor, the biometric information recognition method according to any one of claims 1 to 11 is implemented.

15. A computer program product comprising a computer program or computer executable instructions, characterized in that: When the computer program or computer executable instructions are executed by a processor, the biometric information recognition method according to any one of claims 1 to 11 is implemented.

Citation Information

Patent Citations

  • Providing user video having a virtual curtain to an online conference

    CN105340263A

  • Face image recognition method and device and storage medium

    CN109034013A

  • Image segmentation method, device, and storage medium

    CN109285178A

  • Image processing method and device, electronic equipment and computer readable storage medium

    CN113610884A

  • Background blur processing method, apparatus, and device

    WO2019105262A1