A face detection method and related device

Abstract

The embodiment of the present application provides a face detection method and related equipment, through when detecting that the complete face contour of a user is in the acquisition range, image acquisition is performed on the face of the user, and a face image of the user is obtained; feature extraction is performed on the face image, overall face features and local face features are obtained; the overall face features are compared with overall face reference features, the local face features are compared with local face reference features, and feature difference information is determined; and a face detection result is generated based on the feature difference information. In the embodiment of the present application, the complete face image of the user can be acquired based on the wearable device, and the overall face features and the local face features are extracted from the face image. In the process of detecting the face and generating the face detection result, the overall face features and the local face features are comprehensively considered, and the subjective judgment of the doctor is not needed, so that the detection accuracy can be effectively improved.

Description

Technical Field

[0001] This application relates to the field of data processing, and in particular to a face detection method and related equipment. Background Technology

[0002] Facial features are a core component of "inspection" in traditional Chinese medicine and also one of the reference bases for auxiliary diagnosis in modern medicine.

[0003] Currently, observing a patient's face based on the physician's subjective experience allows for a preliminary assessment of their physical condition. However, because this method relies heavily on the physician's personal experience, the accuracy of current facial detection methods is difficult to guarantee.

[0004] Therefore, there is an urgent need for a solution to address the aforementioned technical problems. Summary of the Invention

[0005] In view of the above problems, this application provides a face detection method and related equipment, aiming to improve the accuracy of face detection.

[0006] The embodiments of this application disclose the following technical solutions:

[0007] First aspect: This application provides a face detection method, including:

[0008] When the user's complete facial contour is detected to be within the acquisition range, the user's face is captured to obtain the user's facial image;

[0009] Feature extraction is performed on the facial image to obtain overall facial features and local facial features;

[0010] The overall facial features are compared with the overall facial reference features, and the local facial features are compared with the local facial reference features to determine the feature difference information.

[0011] Based on the aforementioned feature difference information, facial detection results are generated.

[0012] This application embodiment can acquire a user's complete facial image based on a wearable device and extract overall facial features and local facial features from the facial image. In the process of facial detection and generating facial detection results, both overall facial features and local facial features are considered comprehensively, and there is no need to rely on the subjective judgment of a physician, which can effectively improve the accuracy of detection.

[0013] In one possible implementation, the step of acquiring an image of the user's face when the user's complete facial contour is detected to be within the acquisition range includes:

[0014] When a notification event is triggered on the wearable device, the camera of the wearable device is adjusted to a pre-on state; the wearable device is mounted on the wrist.

[0015] When the wearable device detects that the user's wrist is rotating or hovering, it turns on the camera, which is in a pre-activated state, and when it detects that the user's complete facial contour is within the acquisition range, it acquires an image of the user's face.

[0016] Based on the convenience of wearable devices, this embodiment of the application enables seamless acquisition of facial images when a prompt event is triggered on the wearable device.

[0017] In one possible implementation, the step of acquiring an image of the user's face when the user's complete facial contour is detected to be within the acquisition range includes:

[0018] In response to user-triggered controls on the wearable device, the system detects the user's facial contours. When the system detects that the user's complete facial contours are within the acquisition range, it acquires an image of the user's face.

[0019] In this embodiment of the application, based on the user's operation of any control in the wearable device, the user's face can be captured with the user's authorization, which facilitates subsequent face detection.

[0020] In one possible implementation, the step of acquiring an image of the user's face when the user's complete facial contour is detected to be within the acquisition range includes:

[0021] Based on the user's historical facial detection results, a detection frequency is determined; based on the detection frequency, a detection reminder control is displayed on the user interface of the wearable device; in response to the user triggering the detection reminder control, when the user's complete facial contour is detected within the acquisition range, an image of the user's face is acquired to obtain the user's facial image. In this embodiment, facial detection according to the detection frequency enables long-term detection of the user's face.

[0022] In one possible implementation, the step of acquiring an image of the user's face when the user's complete facial contour is detected to be within the acquisition range includes:

[0023] When the user's complete facial contour is detected within the acquisition range, ambient light is detected;

[0024] When the ambient light intensity is determined to be within the target intensity range, the user's face is captured to obtain the user's facial image;

[0025] When the ambient light intensity is determined to be outside the target intensity range, the supplementary light intensity of the supplementary light module is adjusted based on the ambient light intensity; the user's face is captured to obtain the user's facial image.

[0026] In this embodiment, considering the influence of ambient light on facial images, the intensity of supplementary light can be adjusted based on the supplementary light module to ensure the quality of facial images and thus improve the accuracy of facial detection.

[0027] In one possible implementation, the feature extraction of the facial image to obtain overall facial features and local facial features includes:

[0028] The facial image is subjected to at least one of the following processing methods: grayscale conversion, noise reduction, and smoothing, to generate a preprocessed facial image; feature extraction is performed on the preprocessed facial image to obtain overall facial features and local facial features.

[0029] In this embodiment of the application, preprocessing the facial image can improve the quality of the facial image, thereby improving the accuracy of facial detection.

[0030] In one possible implementation, based on the feature difference information, a face detection result is generated, including:

[0031] Based on the aforementioned feature difference information, a preliminary detection result is generated. The user's historical overall facial features are compared with their current overall facial features, and the user's historical local facial features are compared with their current local facial features to determine the user's facial feature change information. Based on this facial feature change information, the preliminary detection result is corrected to generate a final facial detection result. This avoids detection result bias caused by differences in the user's physical condition at different times, thus improving the accuracy of facial detection.

[0032] In one possible implementation, generating the face detection result based on the feature difference information includes:

[0033] The feature difference information is input into the face detection model to generate face detection results. The face detection model is constructed based on the correlation between the pathogenesis evolution and the feature difference information in the training set.

[0034] In this embodiment, generating facial detection results based on a facial detection model can effectively improve the accuracy of detection.

[0035] In one possible implementation, after generating the face detection result based on the feature difference information, the method further includes:

[0036] Based on the facial detection results, a visual report and / or warning information are generated; the visual report and / or warning information are displayed on the user interface of the wearable device. This can alert the user to abnormalities or allow the user to intuitively view the detection results.

[0037] Second aspect: Embodiments of this application provide a facial detection device, including:

[0038] The system comprises an acquisition unit, an extraction unit, a comparison unit, and a generation unit.

[0039] The acquisition unit is used to acquire an image of the user's face when the complete facial contour of the user is detected to be within the acquisition range.

[0040] The extraction unit is used to extract features from the facial image to obtain overall facial features and local facial features;

[0041] The comparison unit is used to compare the overall facial features with the overall facial reference features, and to compare the local facial features with the local facial reference features to determine feature difference information.

[0042] The generation unit is used to generate face detection results based on the feature difference information.

[0043] Thirdly: This application provides an electronic device, which includes a processor and a memory;

[0044] The memory is used to store program code and transmit the program code to the processor;

[0045] The processor is configured to execute the steps of a face detection method as described in the first aspect above, according to instructions in the program code.

[0046] Fourth aspect: Embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of a face detection method as described in the first aspect above.

[0047] Fifth aspect: This application provides a computer program product, which, when run on a computer, executes the steps of a face detection method as described in the first aspect above.

[0048] Sixth aspect: This application provides a chip including a processor coupled to a memory for executing a computer program or instructions stored in the memory, such that the chip implements the steps of a face detection method as described in the first aspect above. Attached Figure Description

[0049] Figure 1 This is a schematic diagram illustrating an application scenario of a face detection method provided in an embodiment of this application;

[0050] Figure 2 A flowchart illustrating a face detection method provided in this application embodiment;

[0051] Figure 3 This application provides a schematic diagram of the structure of a face detection device according to an embodiment of the present application;

[0052] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0053] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. The terminology used in the following embodiments is for the purpose of describing specific embodiments only and is not intended to be a limitation of this application.

[0054] Currently, observing a patient's face based on the physician's subjective experience allows for a preliminary assessment of their physical condition. However, because this method relies heavily on the physician's personal experience, the accuracy of current facial detection methods is difficult to guarantee.

[0055] Based on this, embodiments of this application provide a face detection method and related equipment. When a user's complete facial contour is detected within the acquisition range, an image of the user's face is acquired to obtain a facial image. Feature extraction is performed on the facial image to obtain overall facial features and local facial features. The overall facial features are compared with overall facial reference features, and the local facial features are compared with local facial reference features to determine feature difference information. Based on the feature difference information, a face detection result is generated.

[0056] This application embodiment can acquire a user's complete facial image based on a wearable device and extract overall facial features and local facial features from the facial image. In the process of facial detection and generating facial detection results, both overall facial features and local facial features are considered comprehensively, and there is no need to rely on the subjective judgment of a physician, which can effectively improve the accuracy of detection.

[0057] The following is combined with Figure 1 The application scenarios provided in the embodiments of this application are introduced, such as... Figure 1 As shown in the figure, this figure is a schematic diagram of an application scenario of a face detection method provided in an embodiment of this application.

[0058] In one example, the method provided in this application embodiment can be executed by a wearable device 1101 and a server 1102, and the wearable device 1101 can interact with the server 1102.

[0059] Wearable devices 1101 may include, but are not limited to, smartwatches, smart bracelets, etc.

[0060] Wearable device 1101 detects the user's facial contour. When the complete facial contour of the user is detected to be within the acquisition range, the device acquires an image of the user's face, obtains the user's facial image, and sends the facial image to server 1102.

[0061] Server 1102 receives a facial image sent by wearable device 1101, extracts features from the facial image to obtain overall facial features and local facial features; compares the overall facial features with overall facial reference features, and compares the local facial features with local facial reference features to determine feature difference information; generates a facial detection result based on the feature difference information, and sends the facial detection result to wearable device 1101.

[0062] After receiving the face detection results sent by the server 1102, the wearable device 1101 can display the face detection results on its user interface for the user to view.

[0063] In another example, the method provided in this application embodiment can be performed by a wearable device.

[0064] Wearable devices detect the user's facial contours. When the complete facial contours of the user are detected within the acquisition range, the device's camera captures an image of the user's face.

[0065] Based on this, the wearable device extracts features from the facial image using its internal software platform to obtain overall facial features and local facial features; it compares the overall facial features with overall facial reference features and the local facial features with local facial reference features to determine feature difference information; and it generates facial detection results based on the feature difference information.

[0066] After a wearable device generates a facial detection result, it can display the result on its user interface for the user to view.

[0067] It should be noted that the user information and data involved in this application (including but not limited to users' facial images, as well as data used for analysis, data stored, data displayed, etc.) are all information and data authorized by the users or fully authorized by all parties, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0068] The following is combined with Figure 2 The face detection method provided in the embodiments of this application will be described, such as... Figure 2 As shown, this figure is a flowchart of a face detection method provided in an embodiment of this application, including S201-S204.

[0069] S201. When the complete facial contour of the user is detected to be within the acquisition range, the user's face is captured to obtain the user's facial image.

[0070] In this embodiment, the user's face can be captured using a wearable device equipped with a camera and a fill light module.

[0071] Considering the impact of facial image quality on facial detection results, this embodiment of the application configures a supplementary lighting module in the wearable device. The supplementary lighting module can automatically adjust the supplementary lighting intensity and angle according to the ambient light intensity to obtain a clear and moderately bright facial image.

[0072] For example, when the wearable device detects that the user's complete facial contours are within the acquisition range, it detects ambient light. If the ambient light intensity is determined to be within a target intensity range, the wearable device's camera can directly capture an image of the user's face to obtain a facial image. If the ambient light intensity is determined to be outside the target intensity range, the wearable device adjusts the supplementary lighting intensity of the supplementary lighting module based on the ambient light intensity, and then captures an image of the user's face using the wearable device's camera to obtain a facial image.

[0073] In the method provided in this application embodiment, with the user's prior authorization, the user's facial image can be captured seamlessly when the user views or operates the wearable device in daily life.

[0074] In one possible implementation, wearable devices are exemplified by devices worn on the wrist, such as smartwatches, smart bracelets, etc.

[0075] Considering that when a reminder event is triggered on a wearable device or when the user views the time, there is a high probability that a facial image of the user that meets the standard can be captured, in this embodiment of the application, when a reminder event is triggered on the wearable device or when the user's wrist is detected to rotate or hover, the camera can be turned on, and when the user's complete facial contour is detected to be within the acquisition range, the user's face image can be acquired to obtain the user's facial image.

[0076] These reminders can include, but are not limited to, sedentary reminders or schedule reminders.

[0077] To improve facial capture efficiency, in this embodiment of the application, when a reminder event is triggered on the wearable device, the camera of the wearable device can be pre-activated. When the wearable device detects that the user's wrist is rotating or hovering, the camera in the pre-activated state is turned on, and when the complete facial contour of the user is detected to be within the capture range, the user's face is captured to obtain the user's facial image.

[0078] On the one hand, turning on a camera in a pre-activated state takes less time than turning it on directly. By first adjusting the wearable device's camera to a pre-activated state, and then turning on the pre-activated camera when the wearable device detects the user's wrist rotation or hovering, the efficiency of facial image acquisition can be improved. On the other hand, compared to the on state, the pre-activated camera consumes less power. In this embodiment, by turning on the pre-activated camera only when the user's wrist rotation or hovering is detected, the wearable device's power consumption can be saved to a certain extent.

[0079] In another possible implementation, the wearable device can detect the user's facial contours in response to the user triggering the operation of controls in the wearable device; when the complete facial contours of the user are detected to be within the acquisition range, the user's face is captured to obtain the user's facial image.

[0080] In one example, the user interface of the wearable device includes a face detection control that is used to trigger the face detection process.

[0081] Users can click the face detection control in the user interface. The wearable device can respond to the user's action by detecting the face detection control and performing facial contour detection. When the complete facial contour of the user is detected within the acquisition range, the device will capture an image of the user's face.

[0082] In another example, with prior user authorization, the wearable device can detect the user's facial contours in response to the user triggering any control on the wearable device; when the user's complete facial contours are detected within the acquisition range, the user's face is captured to obtain the user's facial image.

[0083] For example, when a user taps the screen of a wearable device to turn it on, the wearable device can be activated to detect the user's facial contours; when a user taps a control other than the facial detection control in the wearable device's user interface, the wearable device can be activated to detect the user's facial contours.

[0084] It is understood that, in the embodiments of this application, the control that can trigger the wearable device to detect the user's facial contours is not specifically limited, and it can be set by the user according to their needs.

[0085] In another possible implementation, for long-term detection of the user's face, the method provided in this application embodiment can determine the detection frequency based on the user's historical face detection results; based on the detection frequency, a detection reminder control is displayed on the user interface of the wearable device; in response to the user triggering the detection reminder control, when the user's complete facial contour is detected to be within the acquisition range, the user's face is image acquired to obtain the user's facial image.

[0086] For example, if the user's historical facial detection results are normal, the detection frequency is determined as the first frequency; if the user's historical facial detection results are relatively normal, the detection frequency is determined as the second frequency; if the user's historical facial detection results are relatively abnormal, the detection frequency is determined as the third frequency; and if the user's historical facial detection results are abnormal, the detection frequency is determined as the fourth frequency.

[0087] Among them, the first frequency is less than the second frequency, the second frequency is less than the third frequency, and the third frequency is less than the fourth frequency.

[0088] In another possible implementation, the frequency of personalized detection can be determined based on user information and historical facial detection results. This user information includes, but is not limited to, one or more of the following: user age, gender, basic health status, constitution type (e.g., balanced constitution, yang deficiency constitution, yin deficiency constitution, phlegm-dampness constitution, damp-heat constitution, qi deficiency constitution, blood stasis constitution, qi stagnation constitution, special constitution, etc.), and types of historical or current diseases (e.g., coronary heart disease, diabetes, etc.).

[0089] Based on this, when the wearable device detects that the user's complete facial contour is within the acquisition range, it determines whether the current facial image acquisition meets the personalized detection frequency based on the time interval between the last facial image acquisition and the current time.

[0090] If the facial image acquisition meets the personalized detection frequency, the wearable device can directly and seamlessly acquire the user's facial image. If the facial image acquisition does not meet the personalized detection frequency, a detection reminder control will be displayed on the user interface of the wearable device at the time point that meets the personalized acquisition frequency. In response to the user triggering the detection reminder control, when the user's complete facial contour is detected within the acquisition range, the user's face will be acquired to obtain the user's facial image. That is, the user's face will not be acquired in this instance.

[0091] It is understood that the detection frequency is not specifically limited in the embodiments of this application. For example, the detection frequency can be detection in the morning, noon and evening, three times a day, or once a day or once a week.

[0092] S202. Extract features from the facial image to obtain overall facial features and local facial features.

[0093] In one possible implementation, to improve the quality of the facial image and ensure the accuracy of facial detection, after acquiring the facial image, at least one of the following processes can be performed: grayscale processing, noise reduction processing, and smoothing processing, to generate a preprocessed facial image.

[0094] In this embodiment, the facial image can be sequentially processed by grayscale conversion, noise reduction, and smoothing to generate a preprocessed facial image.

[0095] Among them, grayscale processing is used to reduce the interference of color information in facial images on subsequent processing; denoising processing is used to remove noise points in facial images, making facial images smoother and clearer; smoothing processing is used to enhance the quality of facial images and create favorable conditions for feature extraction.

[0096] Based on this, feature extraction is performed on the preprocessed facial image to obtain overall facial features and local facial features.

[0097] For example, in this embodiment of the application, image segmentation technology can be used to extract the overall facial contour and determine the overall facial region. Within the overall facial region, macroscopic features such as the overall facial gloss, the distribution range and density of pigmentation, and the area and degree of swelling are identified and extracted to obtain overall facial features. These overall facial features can be used to preliminarily determine the user's blood circulation information and organ function information.

[0098] Simultaneously, by extracting microscopic features such as eyelid thickness and color, and lip dryness, moisture, and color intensity within the overall facial area, local facial features can be obtained. These local facial features can be used to provide evidence for associating with corresponding diseases.

[0099] S203. Compare the overall facial features with the overall facial reference features, and compare the local facial features with the local facial reference features to determine the feature difference information.

[0100] In this embodiment, a facial reference feature database can be pre-established, which includes overall facial reference features and local facial reference features. The overall facial reference features are the overall facial features in a healthy state, and the local facial reference features are the local facial features in a healthy state.

[0101] By comparing the overall facial features with overall facial reference features, and by comparing the local facial features with local facial reference features, feature difference information can be determined.

[0102] S204. Based on the feature difference information, generate the face detection result.

[0103] In one possible implementation, a time-series database of facial features can be pre-established, which stores facial images and their corresponding overall and local facial features according to the acquisition time of the facial images.

[0104] For example, overall facial features include, but are not limited to, the location, size, and color value of pigmentation spots; local facial features include, but are not limited to, lip color and eyelid thickness.

[0105] Based on this, preliminary detection results can be generated using the aforementioned feature difference information. The user's historical overall facial features and historical local facial features are retrieved from a time-series database of facial features.

[0106] By comparing the user's historical overall facial features with the current overall facial features, and comparing the user's historical local facial features with the current local facial features, facial feature change information of the user can be determined. Based on the facial feature change information, the preliminary detection results can be corrected to generate facial detection results.

[0107] For example, by using image processing algorithms and statistical methods, facial features corresponding to facial images of the same user collected at different time points can be compared to determine information on changes in facial features.

[0108] This information on changes in facial features may include, but is not limited to, the rate of change and trend of change of overall facial feature parameters or local facial features.

[0109] For example, the rate of change of overall facial feature parameters may include, but is not limited to, the rate of pigmentation diffusion, which can be obtained by dividing the difference between the current pigmentation area and the historical pigmentation area by the time interval; the trend of change may include, but is not limited to, the linear regression trend of lip color changes.

[0110] In one possible implementation, a facial detection model can be pre-constructed based on the correlation between the pathogenesis evolution and the feature difference information in the training set. By inputting the feature difference information into the facial detection model, facial detection results can be generated.

[0111] In the method provided in this application embodiment, after generating the face detection result, a visual report and / or warning information can be generated based on the face detection result, and the visual report and / or warning information can be displayed on the user interface of the wearable device.

[0112] The visualization report may include, but is not limited to, facial feature change trend charts, pathogenesis evolution stage analysis information, health risk warning levels, and personalized intervention suggestions (e.g., adjusting traditional Chinese medicine formulas, adding specific acupoint massage, etc.).

[0113] In summary, the camera and fill light module configured on the wearable device in this embodiment can capture a complete facial image of the user and ensure the image quality of the facial image, thereby ensuring the effectiveness, accuracy and reliability of feature extraction.

[0114] Based on the convenience of wearing wearable devices for extended periods, the method provided in this application embodiment can detect the user's face at a detection frequency or in real time, with the user's authorization, to achieve dynamic tracking of changes in the user's facial features, providing timely and accurate evidence for subsequent diagnosis or disease prevention.

[0115] Meanwhile, the method provided in this application takes into account both overall facial features and local facial features during the process of detecting the face and generating facial detection results, and does not rely on the subjective judgment of doctors, which can effectively improve the accuracy of detection.

[0116] This application provides a facial detection device, see [link to document]. Figure 3 The figure is a schematic diagram of the structure of a face detection device provided in an embodiment of this application. Its specific implementation method is consistent with the implementation method and the technical effect achieved in the embodiments of the above method, and some contents will not be repeated.

[0117] This application provides a facial detection device 3100, including:

[0118] The unit includes a data acquisition unit 3101, an extraction unit 3102, a comparison unit 3103, and a generation unit 3104.

[0119] The acquisition unit 3101 is used to acquire an image of the user's face when the complete facial contour of the user is detected to be within the acquisition range.

[0120] The extraction unit 3102 is used to extract features from the facial image to obtain overall facial features and local facial features;

[0121] The comparison unit 3103 is used to compare the overall facial features with the overall facial reference features, and to compare the local facial features with the local facial reference features to determine feature difference information.

[0122] The generation unit 3104 is used to generate face detection results based on the feature difference information.

[0123] In one possible implementation, the acquisition unit is specifically used for:

[0124] When the user's complete facial contour is detected within the acquisition range, ambient light is detected;

[0125] When the ambient light intensity is determined to be within the target intensity range, the user's face is captured to obtain the user's facial image;

[0126] When the ambient light intensity is determined to be outside the target intensity range, the supplementary light intensity of the supplementary light module is adjusted based on the ambient light intensity; the user's face is captured to obtain the user's facial image.

[0127] In one possible implementation, the acquisition unit is specifically used for:

[0128] When a notification event is triggered on the wearable device, the camera of the wearable device is adjusted to a pre-on state; the wearable device is mounted on the wrist.

[0129] When the wearable device detects that the user's wrist is rotating or hovering, it turns on the camera, which is in a pre-activated state, and when it detects that the user's complete facial contour is within the acquisition range, it acquires an image of the user's face.

[0130] In one possible implementation, the acquisition unit is specifically used for:

[0131] In response to user actions that trigger controls on the wearable device, the system detects the user's facial contours.

[0132] When the user's complete facial contour is detected to be within the acquisition range, the user's face is captured to obtain the user's facial image.

[0133] In one possible implementation, the acquisition unit is specifically used for:

[0134] The detection frequency is determined based on the user's historical facial detection results;

[0135] Based on the detection frequency, a detection reminder control is displayed on the user interface of the wearable device;

[0136] In response to the user triggering the detection reminder control, when the user's complete facial contour is detected to be within the acquisition range, the user's face is captured to obtain the user's facial image.

[0137] In one possible implementation, the extraction unit is specifically used for:

[0138] The facial image is subjected to at least one of the following processes: grayscale conversion, noise reduction, and smoothing, to generate a preprocessed facial image.

[0139] Feature extraction is performed on the preprocessed facial image to obtain overall facial features and local facial features.

[0140] In one possible implementation, the generating unit is specifically used for:

[0141] Based on the aforementioned feature difference information, preliminary detection results are generated;

[0142] The user's historical overall facial features are compared with the current overall facial features, and the user's historical local facial features are compared with the current local facial features to determine the user's facial feature change information;

[0143] Based on the facial feature change information, the preliminary detection results are corrected to generate facial detection results.

[0144] In one possible implementation, the generating unit is specifically used for:

[0145] The feature difference information is input into the face detection model to generate face detection results. The face detection model is constructed based on the correlation between the pathogenesis evolution and the feature difference information in the training set.

[0146] In one possible implementation, the device further includes: a display unit;

[0147] The display unit is specifically used for:

[0148] Based on the facial detection results, a visual report and / or early warning information are generated;

[0149] The visual reports and / or warning information are displayed on the user interface of the wearable device.

[0150] In summary, the camera and fill light module configured on the wearable device in this embodiment can capture a complete facial image of the user and ensure the image quality of the facial image, thereby ensuring the effectiveness, accuracy and reliability of feature extraction.

[0151] Based on the convenience of wearing wearable devices for extended periods, the device provided in this application embodiment can detect the user's face at a detection frequency or in real time, with the user's authorization, to achieve dynamic tracking of changes in the user's facial features, providing timely and accurate evidence for subsequent diagnosis or disease prevention.

[0152] Meanwhile, the device provided in this application embodiment takes into account both overall facial features and local facial features during the process of detecting the face and generating facial detection results, and does not rely on the subjective judgment of doctors, which can effectively improve the accuracy of detection.

[0153] This application provides an electronic device, such as... Figure 4 As shown in the figure, this figure is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device 410 includes: a memory 411 and a processor 412;

[0154] The memory 411 is used to store program code and transmit the program code to the processor 412;

[0155] The processor 412 is used to execute the steps of a face detection method as described above according to the instructions in the program code.

[0156] This application provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of a face detection method as described above.

[0157] This application provides a computer program product that, when run on a computer, executes the steps of a face detection method as described above.

[0158] This application provides a chip including a processor coupled to a memory for executing a computer program or instructions stored in the memory, such that the chip implements the steps of a face detection method as described above.

[0159] The above description is merely one specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A facial detection method, characterized in that, include: In the event of an alert event triggered on a wearable device, and / or in response to a user's action of triggering a control on the wearable device, the user's facial contours are detected; When the user's complete facial contour is detected within the acquisition range, ambient light is detected; When the ambient light intensity is determined to be within the target intensity range, the user's face is captured to obtain the user's facial image; When the ambient light intensity is determined to be outside the target intensity range, the supplementary light intensity of the supplementary light module is adjusted based on the ambient light intensity; the user's face is captured to obtain the user's facial image; Feature extraction is performed on the facial image to obtain overall facial features and local facial features; The overall facial features are compared with the overall facial reference features, and the local facial features are compared with the local facial reference features to determine the feature difference information. Based on the aforementioned feature difference information, preliminary detection results are generated; The user's historical overall facial features are compared with the current overall facial features, and the user's historical local facial features are compared with the current local facial features to determine the user's facial feature change information. Based on the facial feature change information, the preliminary detection results are corrected to generate facial detection results; In the event of a notification event triggered on the wearable device, and / or in response to a user's operation of a control on the wearable device, detecting the user's facial contour includes: When a notification event is triggered on the wearable device, the camera of the wearable device is adjusted to a pre-on state; the wearable device is mounted on the wrist. When the wearable device detects that the user's wrist is rotating or hovering, it turns on the camera, which is in a pre-activated state, to detect the user's facial contours.

2. The method according to claim 1, characterized in that, The process of capturing images of the user's face to obtain the user's facial image includes: The detection frequency is determined based on the user's historical facial detection results; Based on the detection frequency, a detection reminder control is displayed on the user interface of the wearable device; In response to the user triggering the detection reminder control, when the user's complete facial contour is detected to be within the acquisition range, the user's face is captured to obtain the user's facial image.

3. The method according to claim 1, characterized in that, The step of generating facial detection results based on the feature difference information includes: The feature difference information is input into the face detection model to generate face detection results. The face detection model is constructed based on the correlation between the pathogenesis evolution and the feature difference information in the training set.

4. The method according to any one of claims 1-3, characterized in that, After generating the facial detection result based on the feature difference information, the process further includes: Based on the facial detection results, a visual report and / or early warning information are generated; The visual reports and / or warning information are displayed on the user interface of the wearable device.

5. A facial detection device, characterized in that, include: The system comprises an acquisition unit, an extraction unit, a comparison unit, and a generation unit. The acquisition unit is configured to detect the user's facial contour when an alert event is triggered on the wearable device, and / or in response to the user's operation of a control on the wearable device; detect ambient light when the user's complete facial contour is detected within the acquisition range; acquire an image of the user's face when the ambient light intensity is determined to be within a target intensity range; adjust the fill light intensity of the fill light module based on the ambient light intensity when the ambient light intensity is determined to be outside the target intensity range; and acquire an image of the user's face. The extraction unit is used to extract features from the facial image to obtain overall facial features and local facial features; The comparison unit is used to compare the overall facial features with the overall facial reference features, and to compare the local facial features with the local facial reference features to determine feature difference information. The generation unit is used to generate preliminary detection results based on the feature difference information; compare the user's historical overall facial features with the current overall facial features, and compare the user's historical local facial features with the current local facial features to determine the user's facial feature change information; Based on the facial feature change information, the preliminary detection results are corrected to generate facial detection results; The acquisition unit is specifically used to adjust the camera of the wearable device to a pre-on state when a reminder event is triggered on the wearable device; the wearable device is mounted on the wrist. When the wearable device detects that the user's wrist is rotating or hovering, it turns on the camera, which is in a pre-activated state, to detect the user's facial contours.

6. An electronic device, characterized in that, The electronic device includes: a processor and a memory; The memory is used to store program code and transmit the program code to the processor; The processor is configured to execute the steps of a face detection method as described in any one of claims 1-4 according to instructions in the program code.

Images