CN116052251A - Face image optimization method, device, equipment and storage medium - Google Patents

Abstract

The embodiment of the application provides a face image optimization method, a device, equipment and a storage medium, and relates to the technical field of image recognition, wherein the method comprises the following steps: selecting face images recorded in a preset screening period from a face image library as identification images, wherein a plurality of successfully identified face images corresponding to each user are recorded in the face image library; acquiring feature scores corresponding to feature factors of the identification image; acquiring a first index score associated with the identification image according to the feature scores corresponding to the feature factors and the corresponding weight combinations; acquiring the number of times of comparing the identification images to determine a second index score; and determining the recognition rate score of the corresponding recognition image according to the first index score, the second index score and the corresponding evaluation weight, and taking the recognition image with the highest recognition rate score as the preferable image. The scheme realizes the periodical automatic updating of the images in the face image library, and effectively improves the recognition success rate of the system.

Description

Face image optimization method, device, equipment and storage medium

technical field

The embodiments of the present application relate to the technical field of image recognition, and in particular to a face image optimization method, device, equipment, and storage medium.

Background technique

With the development and popularization of face recognition technology, face recognition technology is more and more widely used in business scenarios such as face attendance, face payment, access control, and witness verification. Of course, in a business system that uses face recognition technology, it is necessary to first establish a face image library for system users, and the source of the system's face image library is generally the face image provided by the user when registering, and the business system will use the user's face image database. Save the face images in the database and build a face image database at the same time. When the user uses the face recognition function, the face acquisition device obtains a snapshot image of the user's face, and compares it with the face images in the face image library to determine whether the recognition is successful.

However, as the user grows older, the user's appearance may change, or the use of the scene is different. When the user uses the face recognition function, the angle and light of the face capture device captured by the user will cause the captured image to be different from the face image at the time of registration. The difference is large, resulting in multiple recognition failures when the user performs face recognition, resulting in a high rejection rate and poor user experience.

In related technologies, regular reminders are used to remind users to re-register or update the face images registered in the business system according to relevant requirements, but this method is cumbersome for users, and requires users to be self-disciplined and update pictures within a specified time, so that the user's experience is improved. Poor, and for the business system, the way of self-updating by users is uncontrollable, and it is difficult for the business system to provide effective face recognition functions.

Contents of the invention

The embodiment of the present application provides a face image optimization method, device, equipment and storage medium, which solves the problem of high rejection rate and poor experience caused by the inability to update the image in the library in time, and can effectively screen out high-quality faces. The face image is updated to the face image library without the user's active update, which realizes the periodic automatic update of the image in the face image library, and effectively improves the recognition success rate of the system.

In the first aspect, the embodiment of the present application provides a face image optimization method, the method comprising:

Select the face image recorded in the preset screening period from the face image library, which is recorded with a plurality of successfully identified face images corresponding to each user in the face image library;

Obtain the feature score corresponding to the feature factor of the recognized image;

Acquiring the first index score associated with the recognition image according to the feature score corresponding to the feature factor and the corresponding weight combination;

Obtain the number of comparisons of the recognized image to determine the second index score;

According to the first index score, the second index score and the corresponding evaluation weight, the recognition rate score of the corresponding recognition image is determined, and the recognition image with the highest recognition rate score is selected as the preferred image.

In the second aspect, the embodiment of the present application provides a face image optimization device, including:

The image acquisition module is configured to select the face images recorded in the preset screening period from the face image library as the recognition image, and the face image library is recorded with a plurality of successfully recognized faces corresponding to each user image;

The feature score determination module is configured to obtain the feature score corresponding to the feature factor of the recognition image;

The first score determination module is configured to obtain the first index score associated with the recognition image according to the feature score corresponding to the feature factor and the corresponding weight combination;

The second score determination module is configured to obtain the number of comparisons of the recognized image to determine the second index score;

The image selection module is configured to determine the recognition rate score of the corresponding recognition image according to the first index score, the second index score and the corresponding evaluation weight, and use the recognition image with the highest recognition rate score as the preferred image.

In a third aspect, the embodiment of the present application provides an electronic device, including:

one or more processors;

storage means for storing one or more programs,

When the one or more programs are executed by the one or more processors, the one or more processors implement the face image optimization method described in the embodiment of the first aspect.

In the fourth aspect, the embodiment of the present application further provides a storage medium storing computer-executable instructions, and the computer-executable instructions are used to execute the face image optimization method described in the embodiment of the first aspect when executed by a processor.

This application scores the face images successfully identified in each screening cycle based on the characteristic factors to obtain the corresponding first index score, and also uses each face image as the target image for comparison among the selected face images Carry out recognition comparison, obtain the corresponding second index score, combine the first index score and the second index score, so as to obtain the recognition rate score corresponding to each face image, and then select the preferred image from all face images to add to the human face image. In the face image library, the image in the face image library is automatically updated periodically, which effectively improves the recognition success rate of the system.

Description of drawings

Fig. 1 is the flow chart of the steps of the human face image optimization method provided by the embodiment of the present application;

Fig. 2 is a flow chart of steps for determining the number of times of comparison provided by the embodiment of the present application;

FIG. 3 is a schematic diagram of a human face image optimization device provided in an embodiment of the present application;

FIG. 4 is a schematic structural diagram of an electronic device provided by an embodiment of the present application.

Detailed ways

The embodiments of the present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It can be understood that the specific embodiments described here are only used to explain the embodiments of the present application, but not to limit the embodiments of the present application. In addition, it should be noted that, for the convenience of description, only a part but not all structures related to the embodiment of the present application are shown in the drawings.

The terms "first", "second" and the like in the specification and claims of the present application are used to distinguish similar objects, and are not used to describe a specific sequence or sequence. It should be understood that the terms so used are interchangeable under appropriate circumstances such that the embodiments of the application can be practiced in sequences other than those illustrated or described herein, and that references to "first," "second," etc. distinguish Objects are generally of one type, and the number of objects is not limited. For example, there may be one or more first objects. In the specification and the description of the claims, "plurality" means two or more. In addition, "and/or" in the specification and claims means at least one of the connected objects, and the character "/" generally means that the related objects are an "or" relationship.

The face image optimization method of the present application can be applied to business systems that use face recognition technology, such as business systems that implement business scenarios such as face attendance and face check-in. The face image library, the face image library is pre-stored with face images for face recognition comparison, of course, the captured face images that can be successfully recognized can also be stored in the face image library, it is conceivable that the pre-stored The face image and the captured face image can be distinguished by different identifiers in the face image database. It is conceivable that the service system can be loaded in the electronic device in the form of application software.

Fig. 1 is the flow chart of the steps of the human face image optimization method provided by the embodiment of the present application. As shown in the figure, the human face image optimization method includes the following steps:

Step S110, selecting the face images recorded in the preset screening period from the face image library as the recognition images.

It can be understood that, in the face image library, there are recorded a plurality of successfully recognized face images corresponding to each user, that is, in the face image library, for each registered user in the business system, the business system After each successful recognition, the current captured face images will be stored. For example, the face images corresponding to different users will be identified with the user's system ID (Identity Document, ID card identification number), so that subsequent Screening process.

For the recognition image, the business system selects the face image from the face image library. Of course, the selected face image corresponds to the same user, such as selecting the corresponding face image according to the user's system ID; in addition, the selected face image corresponds to the same user. The face images still correspond to a preset screening cycle. It should be considered that the screening cycle is a preset period of time in the business system, such as one week, one month, etc. During this period, the business system will record the current captured face after each successful recognition. The image is stored and also associated with the current screening cycle.

It should be noted that after the screening of all the identification images in each screening period is completed, that is, after the preferred image corresponding to the screening period is selected, the business system can delete other identification images recorded in the screening period, so as to start a new process. The timing of the screening cycle.

Step S120, acquiring feature scores corresponding to feature factors of the recognized image.

For each recognition image, the business system calculates it to obtain feature scores corresponding to different feature factors. For example, in some embodiments, feature factors include face angle, light brightness, face occlusion, and facial expression. and the proportion of faces, so corresponding to each feature factor, the business system calculates the score of each recognition image separately, so as to obtain the corresponding feature score.

For example, the first feature score corresponding to the angle of the face can be calculated by calculating the face in the face image corresponding to the angle of inclination (such as the Euler angle). The calculation of the angle of inclination can be based on OpenCV and according to the key points of the face Calculate the corresponding Euler angle to determine the corresponding feature score, such as the smaller the tilt angle, the greater the feature score of the item.

For the second characteristic score of light brightness, the corresponding brightness can be calculated according to the RGB components in the face image, for example, the brightness value can be obtained by calculating the RGB three-color light according to the preset proportion.

For the third feature score of face occlusion, it can be calculated by the size of the occlusion range of the face, such as the occlusion range of key parts of the face (such as eyes, nose, mouth and face) on the entire face image The proportion of is used as the characteristic score of this item.

For the fourth feature score of human facial expression, the corresponding facial expression (such as happy, surprised or normal, etc.) can be determined according to the action recognition of facial features (such as eyes, nose, mouth and eyebrows), and for For different facial expressions, different scores can be preset, so as to determine the fourth feature score.

For the fifth feature score of the proportion of faces, the fifth feature score can be determined according to the proportion of faces in the entire image.

It should be noted that for the calculation of the feature scores corresponding to each feature factor, reference may also be made to technical solutions recorded in other existing documents, which will not be described in detail in this embodiment of the present application. In addition, in order to facilitate the calculation of the first index score, the actual value of the feature score can also be standardized to the interval [0, 1].

Step S130, according to the characteristic score corresponding to the characteristic factor and the corresponding weight combination, obtain the first index score associated with the recognition image.

The number of feature weights in the weight combination is the same as the number of feature factors. The feature weights are preset by the business system, and the business system configures corresponding feature weights corresponding to the feature factors, and the sum of all feature weights is 1. The first index score is the value associated with the two, that is, the cumulative sum of the multiplication of all feature factors and their corresponding feature weights.

It should be noted that the first feature weight and the third feature weight take the largest value among all feature weights, that is, the face angle and face occlusion are the largest parameter items in the first index score, which helps to filter out higher quality images.

Step S140, acquiring the matching times of the recognized images to determine the second index score.

For the selected recognition image, it is also necessary to obtain a second index score corresponding to the number of comparisons. It is understandable that after selecting the recognition images recorded in the screening period, it is necessary to count the number of comparisons of each recognition image, for example, compare the images pair by pair, if the similarity between the two images reaches 90% (The threshold can be set according to requirements), then it can be determined that the image is compared, and the number of comparisons is increased by one, thereby traversing all the recognition images, and determining the number of comparisons after each recognition image is compared with all recognition images, so as to obtain each A second index score for the recognition image.

It can be understood that the first index score and the second index score are obtained in no particular order. In some embodiments, the second index score may be calculated first, and then the first index score may be calculated, or simultaneously Calculation of the first index score and the second index score is performed.

Step S150: Determine the recognition rate score of the corresponding recognition image according to the first index score, the second index score and the corresponding evaluation weight, and use the recognition image with the highest recognition rate score as the preferred image.

Similarly, for the evaluation weight, the first evaluation weight is set corresponding to the first index score, and the second evaluation weight is set corresponding to the second index score. The first evaluation weight and the second evaluation weight are also Pre-set in the business system. The recognition rate score is the sum of the product of the first index score, the second index score and the corresponding evaluation weight, so as to obtain the recognition rate score corresponding to each recognition image, and the image with the highest recognition rate score among all recognition images is selected as the preferred image.

It should be noted that, in some embodiments, the first evaluation weight is smaller than the second evaluation weight, so that the second index score accounts for the largest proportion in the recognition rate score, so that the number of comparisons can be used as an important reference for selecting the preferred image indicators, which help to improve the recognition rate.

It can be seen from the above scheme that the business system that applies the face image optimization method provided by this application can calculate the corresponding recognition rate score for the recorded recognition images after each screening cycle, so that the recognition rate in multiple recognition images Select the recognition image with the highest recognition rate as the preferred image, and add it to the face image library, thereby realizing the periodic automatic update of the pictures in the face image library, which can effectively improve the recognition success rate of the system.

Fig. 2 is the flow chart of steps for determining the number of comparisons provided by the embodiment of the present application. As shown in Fig. 2, the face image optimization method also includes the following steps:

Step S210, extracting feature value data of all recognized images.

Step S220 , based on the feature value data, perform a pairwise comparison of all the recognized images, and record the comparison result as the number of comparisons.

It can be understood that, for each recognition image, the extracted feature value data may be recognition data associated with facial features acquired during the face recognition process. After extracting the eigenvalue data of all the recognition images, the recognition images are compared based on the eigenvalue data.

Exemplarily, there are four face images recorded in the screening period, that is, four recognition images, such as recognition image I, recognition image II, recognition image III and recognition image IV. The eigenvalue data are extracted for each recognition image, so as to facilitate the comparison based on the eigenvalue data. The currently selected recognition image I is compared with the recognition image II, recognition image III and recognition image IV respectively. If the similarity of the feature value data of the two compared images is greater than or equal to 90%, it is determined that the ratio of the two images is The number of comparisons is increased by one, so all images are traversed to determine the number of comparisons corresponding to the recognition image I.

For recognition image II, it can be compared with recognition image I, recognition image III, and recognition image IV. It is conceivable that if the previous recognition image I and recognition image II have been compared, then in the current round of comparison For alignment, the last comparison result can be used to reduce the calculation workload of the business system.

Therefore, after comparing the selected identification images, the number of comparisons in each identification image is obtained, so as to obtain the second index score corresponding to each identification image, for example, the ratio of the number of comparisons to the number of comparisons is used as the second index score. indicator score.

It should be noted that, in some embodiments, the identification images themselves are also compared, that is, the number of comparisons and the number of comparisons are cumulatively increased by one.

It can be seen from the above scheme that by comparing the recorded identification images, the business system can determine the number of comparisons of each image, and combine the first index score to screen the images, which helps to select a more representative screening cycle. An image of the results recorded within.

In some embodiments, the screening period is a pre-set duration, and the number of times the user performs face recognition during the screening period is not limited. Therefore, the number of recognition images corresponding to different users recorded during the screening period is different. . In the case that the number of recognized images is too large and the number of recognized images is greater than a preset number, a preset number of recognized images is selected for feature extraction.

For example, when the number of identification images recorded in the screening period is greater than the preset number, the business system can select the identification images satisfying the preset number for screening, that is, select the preferred image from the preset number of identification images. It is conceivable that, for the selection manner of the recognition images, it may be randomly selected, selected according to the recorded order, etc., and the selected number only needs to satisfy the preset number.

It should be noted that, for the situation that the number of identification images recorded in the screening period is less than or equal to the preset number, the business system may select all the identification images recorded in the screening period.

Therefore, by setting the upper limit of the number of selected identification images, the electronic device loaded with the business system can reduce the increased computing workload due to too many identification images recorded during the screening period.

In an application scenario, for example, the business system is applied in the business scenario of face-based time attendance, enterprise employees are used as target users, and the target users are all registered in the business system. In each screening cycle, each time the target user uses the face recognition function to check attendance, the business system stores the currently collected face images in the face image database. If the screening period is set to four weeks in the business system, and the target user checks in at least twice on a working day, corresponding to each target user within the screening period, there are at least 40 successfully recognized faces stored in the face image library. face image.

The business system selects the above-mentioned face image as the recognition image, and calculates the feature score corresponding to each feature factor for each recognition image, such as corresponding to face angle, light brightness, face occlusion, face expression and face Calculate the characteristic score of the five characteristic factors. The business system configures corresponding feature weights for the above feature factors, such as the first feature weight θ ₁ corresponding to the face angle, the second feature weight θ ₂ corresponding to the brightness of light, and the third feature corresponding to face occlusion The weight θ ₃ , the fourth feature weight θ ₄ corresponding to the facial expression, and the fifth feature weight θ ₅ corresponding to the proportion of the face, where θ ₁ , θ ₂ , θ ₃ , θ ₄ and θ ₅ can correspond to The values are 0.3, 0.1, 0.3, 0.1 and 0.2. Therefore, the first index score of the recognized image is calculated, such as calculating the first index score according to the following formula:

Q＝θ ₁ A+θ ₂ B+θ ₃ O+θ ₄ E+θ ₅ P

Among them, Q is the first index score, A is the first feature score corresponding to the face angle, B is the second feature score corresponding to the light brightness, O is the third feature score corresponding to face occlusion, and E is the face expression corresponding P is the fifth feature score corresponding to the face proportion.

Moreover, feature extraction is also performed on each identification image, so as to compare the recorded identification images, so as to obtain the number of comparisons of each identification image, and determine the corresponding second index score T. In addition, according to the first index score Q and the second index score T, the business system is also configured with corresponding evaluation weights, such as the first evaluation weight S 1 corresponding to the first index score Q and the first evaluation weight S ₁ corresponding to the second index score T. Two evaluation weights S ₂ , where S ₁ and S ₂ can take corresponding values. The business system calculates a recognition rate score for each recognition image. For example, the following calculation formula is used to calculate the recognition rate score W:

W＝S ₁ Q+S ₂ T

After obtaining the recognition rate score of each recognition image, the business system can store the recognition image with the highest recognition rate in the face image library, so as to use it as an image for comparison in the face recognition process to realize human The image update of the face image library does not require manual updating by the user. The business system screens out the face image with the highest recognition rate from the face images collected by multiple face recognition functions and automatically updates it to the system's face image library, ensuring The image quality and high image recognition rate improve the user experience effect and effectively improve the recognition success rate of the system.

Fig. 3 is a schematic diagram of the human face image optimization device provided by the embodiment of the present application. The device is used to implement the human face image optimization method provided by the above embodiment, and has corresponding functional modules and beneficial effects for executing the method. As shown in the figure, the device The device includes: an image acquisition module 301 , a feature score determination module 302 , a first score determination module 303 , a second score determination module 304 and an image selection module 305 .

The image acquisition module 301 is configured to select the face images recorded in the preset screening period from the face image library as the recognition image, and the face image library is recorded with a plurality of successfully recognized faces corresponding to each user image;

The feature score determination module 302 is configured to acquire feature scores corresponding to feature factors of the recognized image;

The first score determination module 303 is configured to obtain the first index score associated with the recognition image according to the feature score corresponding to the feature factor and the corresponding weight combination;

The second score determination module 304 is configured to obtain the number of comparisons of the recognized image to determine the second index score;

The image selection module 305 is configured to determine the recognition rate score of the corresponding recognition image according to the first index score, the second index score and the corresponding evaluation weight, and select the recognition image with the highest recognition rate score as the preferred image.

On the basis of the above embodiments, the feature factors include face angle, light brightness, face occlusion, face expression and face proportion.

On the basis of the above embodiments, the weight combination includes feature weights equal to the number of feature factors, and the first index score is the cumulative sum of products of all feature factors and corresponding feature weights.

On the basis of the above-mentioned embodiments, face angle, light brightness, face occlusion, face expression, and face proportion correspond to the first feature weight, the second feature weight, the third feature weight, and the fourth feature in sequence. weight and the fifth feature weight; wherein, the first feature weight and the third feature weight take the largest value among all the feature weights.

On the basis of the above embodiments, there are at least two recognition images, and the second score determination module 304 is also configured to:

extract the eigenvalue data of all recognized images;

Based on the eigenvalue data, pairwise comparisons are performed on all recognized images, and the comparison results are recorded as the number of comparisons.

On the basis of the above embodiments, the second score determination module 304 is further configured to:

When the number of identification images recorded in the screening period is greater than the preset number, the identification images satisfying the preset number are selected.

Based on the above embodiment, the first evaluation weight corresponding to the first index score is smaller than the second evaluation weight corresponding to the second index score.

It is worth noting that, in the embodiment of the above-mentioned human face image optimization device, the various functional modules included are only divided according to the functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be realized; in addition, The specific names of the functional modules are only for the convenience of distinguishing each other, and are not used to limit the protection scope of the present application.

FIG. 4 is a schematic structural diagram of an electronic device provided by an embodiment of the present application. As shown in the figure, the device includes a processor 401, a memory 402, an input device 403, and an output device 404. The number of processors 401 in the device may be one or more, one processor 401 is taken as an example in the figure; the processor 401, memory 402, input device 403 and output device 404 in the device can be connected through a bus or in other ways, and a bus connection is used as an example in the figure. As a computer-readable storage medium, the memory 402 can be used to store software programs, computer-executable programs and modules, such as program instructions/modules corresponding to the face image optimization method in the embodiment of the present application. The processor 401 executes various functional applications and data processing of the device by running the software programs, instructions and modules stored in the memory 402, that is, realizes the above-mentioned face image selection method.

The memory 402 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system and at least one application required by a function; the data storage area can store data created according to the use of the electronic device, such as human The face image, the first index score and the second index score, etc. In addition, the memory 402 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid-state storage devices. In some examples, the memory 402 may further include a memory that is set remotely relative to the processor 401, and these remote memories may be connected to the terminal device through a network. Examples of the aforementioned networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

The input device 403 can be used to receive input numbers or character information, and generate key signal input related to user settings and function control of the device. The output device 404 may be used to send or display key signal output related to user settings and function control of the device, such as outputting a preferred image.

The embodiment of the present application also provides a storage medium storing computer-executable instructions. When executed by a processor, the computer-executable instructions are used to perform related operations in the face image optimization method provided in any embodiment of the present application.

Computer readable storage media includes both volatile and non-permanent, removable and non-removable media implemented by any method or technology for storage of information. Information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read only memory (ROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Flash memory or other memory technology, Compact Disc Read-Only Memory (CD-ROM), Digital Versatile Disc (DVD) or other optical storage, A magnetic tape cartridge, disk storage, or other magnetic storage device, or any other non-transmission medium, that can be used to store information that can be accessed by a computing device.

It should also be noted that the term "comprises", "comprises" or any other variation thereof is intended to cover a non-exclusive inclusion such that a process, method, article, or apparatus comprising a set of elements includes not only those elements, but also includes Other elements not expressly listed, or elements inherent in the process, method, commodity, or apparatus are also included. Without further limitations, an element defined by the phrase "comprising a ..." does not preclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

Note that the above are only preferred embodiments and technical principles used in this application. Those skilled in the art will understand that the present application is not limited to the specific embodiments described herein, and various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the protection scope of the present application. Therefore, although the present application has been described in detail through the above embodiments, the present application is not limited to the above embodiments, and can also include more other equivalent embodiments without departing from the concept of the present application, and the present application The scope is determined by the scope of the appended claims.

Claims (10)

1. A face image optimization method, the method comprising:

selecting face images recorded in a preset screening period from a face image library as identification images, wherein a plurality of successfully identified face images corresponding to each user are recorded in the face image library;

acquiring feature scores corresponding to feature factors of the identification image;

acquiring a first index score associated with the identification image according to the feature scores and the corresponding weight combinations corresponding to the feature factors;

acquiring the number of times of comparison of the identification images to determine a second index score;

and determining the recognition rate score of the corresponding recognition image according to the first index score, the second index score and the corresponding evaluation weight, and taking the recognition image with the highest recognition rate score as a preferable image.

2. The face image optimization method of claim 1 wherein the characteristic factors include face angle, light brightness, face occlusion, face expression, and face duty cycle.

3. The face image optimization method according to claim 1 or 2, wherein the weight combination includes feature weights equal to the number of feature factors, and the first index score is a product cumulative sum of all the feature factors and the corresponding feature weights.

4. The face image optimization method according to claim 2, wherein the face angle, the light brightness, the face occlusion, the face expression, and the face duty ratio correspond to a first feature weight, a second feature weight, a third feature weight, a fourth feature weight, and a fifth feature weight in this order; the first feature weight and the third feature weight take the largest value among all feature weights.

5. The face image optimization method of claim 1, wherein the identification images are at least two, and the obtaining the number of times of comparing the identification images to determine the second index score comprises:

extracting the characteristic value data of all the identification images;

and respectively carrying out pairwise comparison on all the identification images based on the characteristic value data, and recording comparison results as the comparison times.

6. The face image optimization method according to claim 5, further comprising, before extracting the feature value data of all the identification images:

and when the number of the identification images recorded in the screening period is larger than a preset number, selecting the identification images meeting the preset number.

7. The face image preference method of claim 1, wherein a first evaluation weight corresponding to the first index score is smaller than a second evaluation weight corresponding to the second index score.

8. A face image preference apparatus, the apparatus comprising:

the image acquisition module is configured to select face images recorded in a preset screening period from a face image library as identification images, wherein a plurality of face images which are successfully identified and correspond to each user are recorded in the face image library;

the feature score determining module is configured to acquire feature scores corresponding to feature factors of the identification image;

the first score determining module is configured to acquire a first index score associated with the identification image according to the feature scores corresponding to the feature factors and the corresponding weight combinations;

a second score determining module configured to obtain the number of times of comparison of the identification image to determine a second index score;

the image selecting module is configured to determine an identification rate score of the corresponding identification image according to the first index score, the second index score and the corresponding evaluation weight, and take the identification image with the highest identification rate score as a preferred image.

9. An electronic device, comprising:

one or more processors;

storage means for storing one or more programs which when executed by one or more of the processors cause the one or more processors to implement the face image preference method of any one of claims 1 to 7.

10. A storage medium storing computer executable instructions which, when executed by a processor, are adapted to carry out the face image preference method of any one of claims 1 to 7.

Images