Face part classification method, system and device and storage medium
By using external face matching models to preprocess and feature extraction of face image data, and combining sample data sets to determine the category of face parts, the accuracy of face recognition in the makeup field in the prior art is solved, and the user experience improvement of accurate recognition of face shape and facial features and makeup guidance is achieved.
Patent Information
- Application Number
- CN202411880547.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-19
- Publication Date
- 2025-05-06
AI Technical Summary
It is difficult to accurately identify the user's face shape and facial features in the makeup field, and it cannot meet the needs of makeup guidance.
By obtaining face image data and transmitting it to the external face matching model, the face model data of each target part is obtained for preprocessing, feature data is obtained, and the category of each target part is determined based on the sample data set.
It realizes accurate recognition of user face shape and facial features, and improves the user experience guided by makeup.
Smart Images
Figure CN119942608A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of face recognition technology, and in particular to a method, system, device and storage medium for classifying face parts. Background Art
[0002] With the development of computer and Internet industries, face recognition technology is increasingly used in various fields. Among them, the main application is to identify the identity of the user through face recognition. However, this application direction is generally to select individual parts with more obvious features from the facial features as a benchmark, such as face recognition through eyes and nose, so as to identify the identity of the user. However, there are different requirements for face recognition technology in other fields. For example, in the field of makeup, it is more necessary to clearly identify the user's face shape and key facial features to which type, so as to guide the user how to put on makeup, and it does not focus on the user's identity. The current main face recognition solutions that only identify individual facial features do not support the needs of the makeup field, resulting in poor face recognition effects. Summary of the invention
[0003] The present invention provides a method, system, device and storage medium for classifying facial parts to solve one or more technical problems existing in the prior art and at least provide a beneficial choice or create conditions.
[0004] The present invention provides a method for classifying facial parts, the method comprising: Acquire facial image data, and transmit the facial image data to an external face matching model; Obtaining face model data of each target part of the face output by the face matching model, preprocessing the face model data, and obtaining feature data of each target part, wherein the target parts include face shape, eyebrows, eyes, nose and lips; A sample data set is obtained, and a category of each target part is determined and output according to the sample data set and the feature data.
[0005] Optionally, the sample data set includes a sub-data set corresponding to each target part, and the step of determining and outputting the category of each target part according to the sample data set and the feature data includes: For each of the target parts, determining a target sample in the sub-dataset according to the sub-dataset and the feature data, and determining a category corresponding to the target part according to the category of the target sample; Output the categories of all the target parts.
[0006] Optionally, the step of determining a target sample in the sub-dataset according to the sub-dataset and the feature data includes: Calculate the characteristic distances between all sample data in the sub-dataset and the characteristic data respectively; A preset number of the target samples is determined in the sub-dataset according to the characteristic distance, and the characteristic distance of the target sample is the smallest in the sub-dataset.
[0007] Optionally, the step of determining the category corresponding to the target part according to the category of the target sample includes: Voting is performed according to the category of the target sample, and the category of the target part is determined as the category with the largest number of votes.
[0008] Optionally, the step of preprocessing the face model data to obtain feature data of each target part includes: For each of the target part data in the face model data, calculating a standard score for each of the data; When the standard score is greater than a preset threshold, the corresponding data is determined to be abnormal data, and the abnormal data is removed from the face model data to obtain first intermediate data; Performing denoising on the first intermediate data to obtain second intermediate data; The second intermediate data is normalized to obtain the characteristic data.
[0009] Optionally, the step of performing denoising on the first intermediate data to obtain second intermediate data includes: Based on a preset window size, a moving average filtering process is performed on the first intermediate data to obtain the second intermediate data corresponding to each item of the first intermediate data.
[0010] Optionally, the step of normalizing the second intermediate data to obtain the characteristic data includes: Calculate the mean and standard deviation of the second intermediate data, calculate the difference between each item of the second intermediate data and the mean, and calculate the ratio of the difference to the standard deviation, and determine the ratio as the characteristic data corresponding to the second intermediate data.
[0011] In addition, in order to achieve the above purpose, the present application also proposes a face part classification system, the face part classification system comprising: A face acquisition module, which is used to acquire face image data and transmit the face image data to an external face matching model; A data processing module, the data processing module is used to obtain face model data of each target part of the face output by the face matching model, pre-process the face model data, and obtain feature data of each target part, the target parts including face shape, eyebrows, eyes, nose and lips; An algorithm analysis module is used to obtain a sample data set, and determine and output a category of each target part according to the sample data set and the feature data.
[0012] In addition, in order to achieve the above-mentioned purpose, the present application also proposes a facial part classification device, which includes: a memory, a processor, and a facial part classification program stored in the memory and executable on the processor, wherein the facial part classification program is configured to implement the steps of the facial part classification method as described in any of the above items.
[0013] In addition, in order to achieve the above-mentioned purpose, the present application also proposes a storage medium, on which a facial part classification program is stored, and when the facial part classification program is executed by a processor, the steps of the facial part classification method described in any of the above items are implemented.
[0014] The present invention has at least the following beneficial effects: by using an external face matching model to convert the collected user's face image data into face model data, and classifying the face model data according to the target parts, thereby obtaining data of several target parts that need to be analyzed, such as face shape, eyebrows, eyes, nose and lips, based on which, the face model data is preprocessed to obtain feature data, and the feature data is analyzed in combination with the sample data set to determine the category of each target part. Compared with conventional face recognition technology that only identifies individual facial features and aims to identify the identity of the user, the present invention selects several target parts for key analysis and extracts the corresponding face model data, based on which, the category of each target part is analyzed and identified, so as to realize the identification of the category of each target part of the user, improve the accuracy of identifying the user's face shape and facial features, and also help to improve the user experience of makeup guidance. BRIEF DESCRIPTION OF THE DRAWINGS
[0015] The accompanying drawings are used to provide a further understanding of the technical solution of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the technical solution of the present invention and do not constitute a limitation on the technical solution of the present invention.
[0016] Figure 1 It is a flowchart of a first embodiment of a method for classifying facial parts according to the present invention; Figure 2 It is a flowchart of a second embodiment of the method for classifying facial parts of the present invention; Figure 3 It is a flowchart of a third embodiment of the method for classifying facial parts of the present invention; Figure 4 It is a structural schematic diagram of the face part classification system of the present invention; Figure 5The figure is a schematic diagram of the structure of the device involved in the operation of an embodiment of the face part classification device of the present invention. DETAILED DESCRIPTION
[0017] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.
[0018] It should be noted that, although the functional modules are divided in the system schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the system or the order in the flowchart. The terms "first", "second", etc. in the specification, claims and the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.
[0019] The embodiment of the present invention provides a method for classifying facial parts. Figure 1 , a first embodiment of the face part classification method of the present applicant is proposed. In this embodiment, the face part classification method includes: Step S10, acquiring face image data, and transmitting the face image data to an external face matching model.
[0020] The facial image data is an image including the user's face. The image can be obtained by the user controlling a mobile terminal or other shooting device. After uploading the image through a preset interface, the facial image data of the user is obtained. In this embodiment, the facial image data uploaded by each user is received through the preset interface. At this time, the facial image data can be preliminarily analyzed to determine whether the image includes a face to prevent the user from accidentally uploading images other than those used for facial recognition. After determining that there is no problem with the image, the facial image data is transmitted to an external face matching model.
[0021] Step S20, obtaining face model data of each target part of the face output by the face matching model, preprocessing the face model data, and obtaining feature data of each target part, the target parts including face shape, eyebrows, eyes, nose and lips.
[0022] The external face matching model is the FaceMesh model of Mediapipe, which can recognize faces in the input image and convert them into three-dimensional feature point data. In this embodiment, the number of feature points is 468, and the data output by the face matching model is defined as face model data. By selecting the corresponding face parts, the face model data corresponding to each target part can also be determined.
[0023] The target parts include face shape, eyebrows, eyes, nose and lips. The above five target parts are the parts that need to be analyzed in the field of makeup. They are suitable for different makeup styles or combinations according to their categories and have a greater impact on the final makeup effect. In other embodiments, other parts of the face may also be included.
[0024] Before using face model data for analysis, the face model data needs to be preprocessed. The preprocessing steps include data cleaning, denoising and normalization to highlight the features in the data and improve the accuracy of the analysis.
[0025] Step S30, obtaining a sample data set, determining and outputting the category of each target part according to the sample data set and the feature data.
[0026] It should be noted that the category of the target part described in this embodiment does not refer to the category of distinguishing different target parts, such as distinguishing eyebrows and eyes, but refers to the specific category of each target part, such as the eyebrow shape category, the eye shape category and the nose shape category.
[0027] Specifically, a sample data set is obtained, and the sample data set includes a corresponding sub-data set corresponding to each target part. The sub-data set includes multiple sample data of various categories of the target part. Each sample data has corresponding annotation information to determine the category of the sample data. Based on this, for the feature data of each target category, the category of each target part can be determined by comparing the corresponding sub-data set and the feature data, and then the category of each target part is output.
[0028] The embodiments of the present invention have at least the following beneficial effects: by using an external face matching model to convert the collected user's face image data into face model data, and classifying the face model data according to the target parts, thereby obtaining data of several target parts that need to be analyzed, such as face shape, eyebrows, eyes, nose and lips, based on which, the face model data is preprocessed to obtain feature data, and the feature data is analyzed in combination with the sample data set to determine the category of each target part. Compared with conventional face recognition technology that only identifies individual facial features and aims to identify the identity of the user, the present invention selects several target parts for key analysis and extracts the corresponding face model data, based on which, the category of each target part is analyzed and identified, so as to realize the identification of the category of each target part of the user, improve the accuracy of identifying the user's face shape and facial features, and also help to improve the user experience of makeup guidance.
[0029] Further, based on the above embodiment, a second embodiment of the face part classification method of the present applicant is proposed. Figure 2 In this embodiment, the step of determining and outputting the category of each target part according to the sample data set and the feature data includes: Step S31 : for each target part, a target sample in the sub-data set is determined according to the sub-data set and the feature data, and a category of the corresponding target part is determined according to the category of the target sample.
[0030] Step S32, output the categories of all target parts.
[0031] Specifically, for each target part, the data of the target part in the feature data is determined, and the sub-dataset of the target part in the sample data set is determined. Based on the determined feature data and sub-dataset, the target sample closest to the feature data is determined from the sub-dataset, and the category of the corresponding target part is judged by the category marked by the target sample.
[0032] By combining the target samples to determine the category of the user's target part, the classification and identification of the target part is based on comparison and analysis with other real sample references, rather than purely based on numerical indicators. This improves the accuracy of identifying the target part category, thereby improving the accuracy of identifying the user's face shape and facial features.
[0033] Furthermore, in this embodiment, the step of determining the target sample in the sub-data set according to the sub-data set and the feature data includes: Calculate the characteristic distance between all sample data in the sub-dataset and the characteristic data.
[0034] A preset number of target samples are determined in the sub-dataset according to the characteristic distance, and the characteristic distance of the target samples is the smallest in the sub-dataset.
[0035] Specifically, first determine the preset number of selected target samples, for example, 5, and calculate the characteristic distance between each sample data and the characteristic data in the sub-dataset. The characteristic distance is one of the Euclidean distance, Manhattan distance or Minkowski distance. Taking the Euclidean distance as an example, the calculation formula of the Euclidean distance is: The smaller the feature distance, the more similar the sample data is to the feature data, and the larger the feature distance, the more dissimilar the sample data is to the feature data. A preset number of sample data with the smallest feature distance are selected from the sample data according to the feature data and determined as the target samples.
[0036] In other embodiments, in order to avoid dissimilar sample data from being selected as target samples, it can be determined whether the characteristic distance of each target sample is less than or equal to a preset distance threshold. When it is determined to be less than or equal to, the target sample is normal. When it is determined to be greater than, the target sample is eliminated to prevent interference with the classification of subsequent target parts.
[0037] By calculating the feature distance between the sample data and the feature data, the target sample is determined according to the feature distance, and the target sample is determined from the perspective of similarity, thereby improving the accuracy of the target sample, thereby improving the accuracy of identifying the target part category and the user's face shape and facial features.
[0038] Furthermore, in this embodiment, the step of determining the category of the corresponding target part according to the category of the target sample includes: Voting is performed according to the category of the target sample, and the category of the target part is determined to be the category with the most votes.
[0039] For the determined target sample, voting statistics are performed according to the category of the target sample to determine the category with the most votes. For example, the target sample includes 3 votes for nose type A and 2 votes for nose type B, among which the category with the most votes is nose type A. Based on this, the category of the target part is determined.
[0040] In addition, if there are two or more categories with the same number of votes, the two or more categories can be output and manually selected and judged by the user.
[0041] The category of the target sample is analyzed through a voting mechanism, and the judgment result of the category of the target part is determined, so that the category of the target part can be intelligently identified based on the feature data and sample data set, thereby improving the accuracy of identifying the user's face shape and facial features.
[0042] In other embodiments, after outputting the category of the target part, it is also possible to determine whether to add the feature data to the sample data set based on user feedback on recognition accuracy and privacy requirements, thereby further expanding the sample data of the sample data set and improving the recognition accuracy of the target part category.
[0043] Further, based on the above embodiments, a third embodiment of the face part classification method of the present applicant is proposed. Figure 3 In this embodiment, the step of preprocessing the face model data to obtain feature data of each target part includes: Step S21, for each target part of the face model data, calculate the standard score of each item of data.
[0044] Step S22: when the standard score is greater than a preset threshold, the corresponding data is determined to be abnormal data, and the abnormal data is removed from the face model data to obtain first intermediate data.
[0045] Step S23, denoising the first intermediate data to obtain second intermediate data.
[0046] Step S24, normalizing the second intermediate data to obtain feature data.
[0047] Specifically, the preprocessing steps include data cleaning, denoising and normalization processing, wherein the three processing steps are performed in order, and in order to avoid interference between the data of different target parts, the preprocessing steps are performed for the data of each target part respectively. In this embodiment, the processing of the data of a single target part is used as an example for explanation, and no further description is given. First, the standard score of each data in the face model data corresponding to the target part is calculated, and the calculation formula of the standard score is: ,in, is the standard score, is the value of the data. is the average value, is the standard deviation, and at the same time, the preset threshold of the target part is determined, and the standard score calculated for each data is compared with the preset threshold. When the standard score is less than or equal to the preset threshold, the data is normal, and when the standard score is greater than the preset threshold, the data is judged as outlier abnormal data, and the abnormal data is removed from the face model data to complete the data cleaning step, and the data after the data cleaning step is defined as the first intermediate data. Based on this, the denoising step is further performed to obtain the second intermediate data, and then the normalization step is performed to obtain the feature data.
[0048] By calculating the standard score, we can find abnormal data in the face model data and complete data cleaning to avoid abnormal data interfering with the category recognition of the target part and affecting the recognition results, thereby improving the accuracy of identifying the category of the target part and thus improving the accuracy of identifying the user's face shape and facial features.
[0049] Furthermore, in this embodiment, the step of performing denoising on the first intermediate data to obtain the second intermediate data includes: Based on a preset window size, a moving average filtering process is performed on the first intermediate data to obtain second intermediate data corresponding to each item of the first intermediate data.
[0050] A preset window size, such as 3 or 5, is obtained, and a moving average filtering process is performed on the first intermediate data based on the preset window size, and the calculation formula is: ,in, is the moving average, is the data item number of the first intermediate data, To preset the window size, In order to calculate the data item number in the window, the moving average of each data item in the first intermediate data is calculated, and for the starting data item and the ending data item in the sequence, the data items can be filled by zero filling or periodic filling, so as to calculate their moving average value, and the calculated moving average value forms the second intermediate data.
[0051] By calculating the moving average, the first intermediate data is filtered and denoised, thereby reducing the data noise that may exist in the feature data, improving the accuracy of identifying the target part category, and thus improving the accuracy of identifying the user's face shape and facial features.
[0052] Furthermore, in this embodiment, the step of normalizing the second intermediate data to obtain feature data includes: The mean and standard deviation of the second intermediate data are calculated, the difference between each item of the second intermediate data and the mean is calculated, and the ratio of the difference to the standard deviation is calculated, and the ratio is determined as the characteristic data corresponding to the second intermediate data.
[0053] Specifically, the calculation formula in the normalization processing step can refer to the calculation formula in the above-mentioned data cleaning step. However, the calculation formula is used to calculate the standard score in the data cleaning step, and the outlier abnormal data in the face model data is found according to the standard score. The standard score is not used in subsequent steps. In the normalization processing step, the normalized value obtained by the calculation formula is used as the value of the feature data. Although the same design of the calculation formula is used, the role it plays in the two steps is not the same. The normalized value of each data in the second intermediate data is calculated to form feature data.
[0054] By obtaining feature data through normalization processing, the face model data is finally standardized, which facilitates the analysis of the face model data from the same scale, improves the accuracy of identifying the target part category, and thus improves the accuracy of identifying the user's face shape and facial features.
[0055] In addition, the embodiment of the present invention also provides a face part classification system, referring to Figure 4 ,The face part classification system includes: face acquisition module, data processing module and algorithm analysis module. Among them: The face acquisition module is used to obtain face image data and transmit the face image data to an external face matching model.
[0056] The data processing module is used to obtain the face model data of each target part of the face output by the face matching model, pre-process the face model data, and obtain the feature data of each target part, the target parts including face shape, eyebrows, eyes, nose and lips.
[0057] The algorithm analysis module is used to obtain a sample data set, and determine and output the category of each target part based on the sample data set and feature data.
[0058] The contents of the above method embodiments are all applicable to the present system embodiments. The functions specifically implemented by the present system embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.
[0059] On the other hand, reference Figure 5 , Figure 5 It is a schematic diagram of the device structure of a face part classification device.
[0060] The embodiment of the present invention further provides a face part classification device, comprising: a processor and a memory, wherein the memory is used to store a computer readable program. When the computer readable program is executed by the processor, the processor implements the face part classification method as described in any one of the above technical solutions.
[0061] It will be appreciated by those skilled in the art that all or some of the steps and systems in the methods disclosed above may be implemented as software, firmware, hardware, and appropriate combinations thereof. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or temporary medium). As known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by a computer. As is well known to those skilled in the art, communication media typically embodies computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.
[0062] In addition, an embodiment of the present invention further provides a storage medium, on which a facial part classification program is stored. When the facial part classification program is executed by a processor, the relevant steps of any embodiment of the facial part classification method described above are implemented.
[0063] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein, for example. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0064] It should be understood that in the present application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the objects associated before and after are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.
[0065] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.
[0066] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0067] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.
[0068] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or all or part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, referred to as ROM), random access memory (Random Access Memory, referred to as RAM), disk or optical disk and other media that can store program codes.
[0069] Although the description of the present application has been quite detailed and specifically describes several embodiments, it is not intended to be limited to any of these details or embodiments or any particular embodiment, but should be regarded as providing a broad possible interpretation of these claims by reference to the attached claims, taking into account the prior art, so as to effectively cover the intended scope of the present application. In addition, the above description of the present application is based on the embodiments foreseeable by the inventor, and its purpose is to provide a useful description, and those non-substantial changes to the present application that have not yet been foreseen may still represent equivalent changes to the present application.
Claims
1. A method for classifying facial parts, characterized in that: The face part classification method comprises: Acquire facial image data, and transmit the facial image data to an external face matching model; Obtaining face model data of each target part of the face output by the face matching model, preprocessing the face model data, and obtaining feature data of each target part, wherein the target parts include face shape, eyebrows, eyes, nose and lips; A sample data set is obtained, and a category of each target part is determined and output according to the sample data set and the feature data.
2. The method for classifying facial parts according to claim 1, characterized in that: The sample data set includes a sub-data set corresponding to each target part, and the step of determining and outputting the category of each target part according to the sample data set and the feature data includes: For each of the target parts, determining a target sample in the sub-dataset according to the sub-dataset and the feature data, and determining a category corresponding to the target part according to the category of the target sample; Output the categories of all the target parts.
3. The method for classifying facial parts according to claim 2, characterized in that: The step of determining the target sample in the sub-dataset according to the sub-dataset and the feature data comprises: Calculate the characteristic distances between all sample data in the sub-dataset and the characteristic data respectively; A preset number of the target samples is determined in the sub-dataset according to the characteristic distance, and the characteristic distance of the target sample is the smallest in the sub-dataset.
4. The method for classifying facial parts according to claim 2, characterized in that: The step of determining the category corresponding to the target part according to the category of the target sample comprises: Voting is performed according to the category of the target sample, and the category of the target part is determined as the category with the largest number of votes.
5. The method for classifying facial parts according to claim 1, characterized in that: The step of preprocessing the face model data to obtain feature data of each target part comprises: For each of the target part data in the face model data, calculating a standard score for each of the data; When the standard score is greater than a preset threshold, the corresponding data is determined to be abnormal data, and the abnormal data is removed from the face model data to obtain first intermediate data; Performing denoising on the first intermediate data to obtain second intermediate data; The second intermediate data is normalized to obtain the characteristic data.
6. The method for classifying facial parts according to claim 5, characterized in that: The step of performing denoising on the first intermediate data to obtain second intermediate data comprises: Based on a preset window size, a moving average filtering process is performed on the first intermediate data to obtain the second intermediate data corresponding to each item of the first intermediate data.
7. The method for classifying facial parts according to claim 5, characterized in that: The step of normalizing the second intermediate data to obtain the characteristic data comprises: Calculate the mean and standard deviation of the second intermediate data, calculate the difference between each item of the second intermediate data and the mean, and calculate the ratio of the difference to the standard deviation, and determine the ratio as the characteristic data corresponding to the second intermediate data.
8. A facial part classification system, characterized in that: The face part classification system comprises: A face acquisition module, which is used to acquire face image data and transmit the face image data to an external face matching model; A data processing module, the data processing module is used to obtain face model data of each target part of the face output by the face matching model, pre-process the face model data, and obtain feature data of each target part, the target parts including face shape, eyebrows, eyes, nose and lips; An algorithm analysis module is used to obtain a sample data set, and determine and output a category of each target part according to the sample data set and the feature data.
9. A facial part classification device, characterized in that: The face part classification device comprises: a memory, a processor and a face part classification program stored in the memory and executable on the processor, wherein the face part classification program is configured to implement the steps of the face part classification method as described in any one of claims 1 to 7.
10. A storage medium, characterized in that: The storage medium stores a facial part classification program, which, when executed by the processor, implements the steps of the facial part classification method according to any one of claims 1 to 7.