Beauty instrument control method, skin data processing method, control system, and device

Abstract

Provided are a beauty instrument control method, a skin data processing method, a control system, and a device. The present application relates to the field of biological medical treatment. The method comprises: by means of a touch chip in a beauty instrument, determining whether an end surface of the beauty instrument fits the skin; if the end surface of the beauty instrument fits the skin, starting a beauty instrument load, and on the basis of motion estimation technology and by means of a camera of the beauty instrument, detecting whether a relative movement occurs between the skin and the end surface of the beauty instrument; if no relative movement occurs between the skin and the end surface of the beauty instrument, reducing the power of the beauty instrument load until the beauty instrument is in a standby state, and re-executing the step of determining, by means of the touch chip in the beauty instrument, whether the end surface of the beauty instrument fits the skin; and if a relative movement occurs between the skin and the end surface of the beauty instrument, controlling the beauty instrument to work normally. Further disclosed in the present application are the corresponding skin data processing method, the control system and the device.

Description

Beauty instrument control method, skin data processing method, control system and equipment

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority to Chinese application CN202311037371.2 filed on August 17, 2023, Chinese application CN202311045825.0 filed on August 17, 2023, Chinese application CN202410275233.6 filed on March 12, 2024, and Chinese application CN202410325752.9 filed on March 21, 2024, the entire contents of which are incorporated by reference into this application. Technical Field

[0003] The present application relates to the field of biomedicine, and in particular to a beauty instrument control method, skin data processing method, control system and equipment. Background Art

[0004] At present, most beauty devices may have the following problems during use:

[0005] (1) In actual applications, there are situations where the beauty instrument and the skin are relatively stationary, while the person and the instrument are in relative motion with respect to the ground. However, when using touch sensing technology and motion sensing technology commonly used in existing technologies to obtain the relative state between the front end of the device and the skin, the current relative motion state between the front end of the device and the skin cannot be identified, thus greatly reducing the safety of the device output energy control.

[0006] (2) Most beauty devices use a simple method of automatic timing of liquid discharge, which may lead to problems such as uncertain automatic liquid discharge and fixed liquid discharge volume, incorrect liquid discharge location and time, and the inability to control the liquid at a fixed point and quantity.

[0007] (3) In beauty scenarios where beauty devices are automatically controlled, beauty devices can generally only perform uniform operations on the entire face according to fixed preset working parameters, and cannot achieve precise skin care.

[0008] (4) During the use of milk light laser beauty equipment, the temperature of the instrument needs to be monitored to avoid excessive temperature causing burns to the user or damage to the instrument. The existing conventional heat dissipation technology mostly uses single or integrated fusion technologies such as heat pipes, TEC (Thermoelectric Cooler, semiconductor radiator) and air cooling in parallel. However, this heat dissipation technology not only increases the size of the equipment, but also cannot achieve high heat dissipation efficiency due to the particularity of the application scenario. Ordinary TEC heat dissipation has the problem that the temperature of the heat dissipation module close to the tungsten filament will also rise due to the high temperature of the tungsten filament light and heat radiation itself. At this time, the temperature difference between the preset target temperature and the actual heat dissipation module temperature is too large. The larger the TEC temperature difference, the worse the cooling effect, and the working efficiency of the load TEC decreases accordingly.

[0009] Application Contents

[0010] In view of this, some embodiments of the present application can provide a beauty instrument control method, a skin data processing method, a control system and a device, in an effort to alleviate at least one of the above problems.

[0011] The present application provides a beauty instrument control method, the method comprising:

[0012] The touch chip of the beauty instrument is used to identify whether the end face of the beauty instrument is in contact with the skin;

[0013] If the end face of the beauty instrument is in contact with the skin, the beauty instrument load is started, and the camera of the beauty instrument is used to detect whether the skin and the end face of the beauty instrument are in relative motion based on motion estimation technology;

[0014] If the skin and the end face of the beauty instrument do not move relative to each other, the power of the beauty instrument load is reduced until the beauty instrument is in a standby state; and the step of identifying whether the end face of the beauty instrument is in contact with the skin by using the touch chip of the beauty instrument is repeated;

[0015] If the skin and the end surface of the beauty instrument move relative to each other, the beauty instrument is controlled to operate normally.

[0016] The present application provides a beauty instrument control device, the device comprising:

[0017] an identification module configured to identify whether the end face of the beauty instrument is in contact with the skin through a touch chip of the beauty instrument;

[0018] a detection module configured to activate a beauty instrument load if the end face of the beauty instrument is in contact with the skin, and detect whether relative motion occurs between the skin and the end face of the beauty instrument using a camera based on motion estimation technology;

[0019] The first control module is configured to reduce the power of the load until the beauty instrument is in a standby state if there is no relative movement between the skin and the end face of the beauty instrument; and re-execute the step of identifying whether the end face of the beauty instrument is in contact with the skin through the touch chip of the beauty instrument;

[0020] The second control module is configured to control the beauty instrument to operate normally if the skin and the end surface of the beauty instrument move relative to each other.

[0021] The present application provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the steps of any of the above-described beauty instrument control methods are implemented.

[0022] The present application provides a computer-readable storage medium storing a computer program, which implements the steps of any of the above beauty instrument control methods when executed by a processor.

[0023] The present application provides a skin data processing method, which should be configured as a skin beauty device with a high-speed multispectral imaging module at the front end; the skin data processing method includes:

[0024] Acquiring a current detailed skin image acquired by the high-speed multispectral imaging module;

[0025] Determining the current skin position of the skin beauty device by comparing the current detailed skin image with a pre-collected full-face skin image;

[0026] Determine current treatment parameters of the skin beauty device according to the current detailed skin image and the current skin position.

[0027] The present application provides a skin data processing device, which should be configured as a skin beauty instrument with a high-speed multispectral imaging module at the front end; the skin data processing device includes:

[0028] an image acquisition module configured to acquire the current detailed skin image acquired by the high-speed multispectral imaging module;

[0029] a position determination module configured to determine a current skin position of the skin beauty device by comparing the current detailed skin image with a pre-collected full-face skin image;

[0030] The parameter determination module is configured to determine the current treatment parameters of the skin beauty device according to the current detailed skin image and the current skin position.

[0031] The present application provides a skin beauty instrument, comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and when the processor executes the computer program, it implements any of the skin data processing methods described above.

[0032] The present application provides a skin beauty system, comprising the skin beauty instrument described above, and also comprising a user terminal and a cloud server respectively connected to the skin beauty instrument.

[0033] This application provides a multi-dimensional control method for a beauty device, which should be configured as a server and includes:

[0034] Obtaining electrode signals, skin sampling data, and image data collected by beauty equipment;

[0035] determining whether the beauty device is in contact with the skin of the current user based on the electrode signal;

[0036] If the skin of the current user is in contact, determining the working state of the beauty device based on the skin sampling data;

[0037] The skin condition of the current user is determined based on a pre-trained skin detection model and the image data.

[0038] The present application provides a multi-dimensional control device for a beauty device, which should be configured as a server and includes:

[0039] a data acquisition module configured to acquire electrode signals, skin sampling data, and image data collected by the beauty device;

[0040] an electrode signal determination module, configured to determine whether the beauty device is in contact with the skin of a current user based on the electrode signal;

[0041] a working state determining module configured to determine the working state of the beauty device based on the skin sampling data if the skin of the current user is in contact;

[0042] The skin condition determination module determines the skin condition of the current user based on a pre-trained skin detection model and the image data.

[0043] The present application provides a beauty system, comprising: a beauty device, and a server connected to the beauty device.

[0044] The present application provides a photon beauty device control system, which is applied to a photon beauty device. The system includes a first processor, a temperature adjustment module, and an alarm module. The temperature adjustment module includes a semiconductor refrigeration unit, a water cooling unit, a radiator, and a fan.

[0045] The first processor is communicatively connected to the temperature adjustment module, and the first processor is electrically connected to the alarm module;

[0046] The first processor is also electrically connected to a load, and when the load is in a working state, the first processor controls the temperature adjustment module to operate;

[0047] The first processor is configured to increase the power of the semiconductor refrigeration unit, the water cooling unit, and the fan when the temperature adjustment module detects that the instrument temperature of the photonic beauty device exceeds a temperature threshold;

[0048] The first processor is further configured to, if the temperature of the instrument does not fall back to the temperature threshold within a first preset time period, reduce the power of the load until the temperature of the instrument falls back to the temperature threshold;

[0049] If the temperature of the instrument does not fall back to the temperature threshold within a second preset time period after the load power is reduced, the load is controlled to shut down and the alarm module issues an alarm.

[0050] The present application provides a photon cosmetic device, which includes the photon cosmetic device control system described in any one of the above items.

[0051] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without creative work.

[0053] FIG1 shows a flow chart of a beauty instrument control method provided by an embodiment of the present application;

[0054] FIG2 is a schematic diagram showing the positional relationship between a beauty instrument and a camera provided in an embodiment of the present application;

[0055] FIG3 shows another flow chart of a beauty instrument control method provided by an embodiment of the present application;

[0056] FIG4 shows a schematic structural diagram of a beauty instrument control device provided in an embodiment of the present application;

[0057] FIG5 is a schematic flow chart showing a skin data processing method provided in an embodiment of the present application;

[0058] FIG6 is a schematic flow chart showing another skin data processing method provided in an embodiment of the present application;

[0059] FIG7 shows a schematic structural diagram of a skin data processing device provided in an embodiment of the present application;

[0060] FIG8 shows a schematic structural diagram of a skin beauty device provided in an embodiment of the present application;

[0061] FIG9 shows a schematic structural diagram of a skin care system provided in an embodiment of the present application;

[0062] FIG10 shows a flow chart of a multi-dimensional control method for a beauty device provided by an embodiment of the present application;

[0063] FIG11 shows a training method for a skin detection model provided in an embodiment of the present application;

[0064] FIG12 shows a schematic diagram of a skin detection model provided in an embodiment of the present application;

[0065] FIG13 shows a schematic diagram of an encoder network structure provided in an embodiment of the present application;

[0066] FIG14 shows a schematic diagram of a lightweight skin classification model provided in an embodiment of the present application;

[0067] FIG15 shows a schematic structural diagram of a Vit sub-network provided in an embodiment of the present application;

[0068] FIG16 shows a schematic diagram of the structure of a self-attention sub-network provided in an embodiment of the present application;

[0069] FIG17 shows a schematic structural diagram of a multi-dimensional control device for a beauty device provided in an embodiment of the present application;

[0070] FIG18 shows a schematic structural diagram of a beauty device provided in an embodiment of the present application.

[0071] FIG19 shows a schematic diagram of a module of a photonic beauty device control system provided in an embodiment of the present application;

[0072] FIG20 shows a diagram of the MCU pin definition of a photonic beauty device control circuit provided in an embodiment of the present application;

[0073] FIG21 shows a schematic diagram of a TEC control circuit of a photonic beauty device control circuit provided in an embodiment of the present application;

[0074] FIG22 shows a schematic diagram of an NTC control circuit of a photonic beauty device control circuit provided in an embodiment of the present application;

[0075] FIG23 shows a power supply diagram of a specific implementation of a photonic beauty device control system provided in an embodiment of the present application; and

[0076] FIG24 shows a power supply diagram of another specific implementation of the photonic beauty device control system provided in an embodiment of the present application.

[0077] Icons: 201-camera; 202-beauty instrument end face; 401-identification module; 402-detection module; 403-first control module; 404-second control module; 100-first processor; 200-temperature adjustment module; 300-alarm module; 400-load; 500-vibration motor; 600-second processor; 700-display screen; 800-camera; 210-semiconductor refrigeration unit; 220-water cooling unit; 230-radiator; 240-fan. DETAILED DESCRIPTION

[0078] In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work are within the scope of protection of this application.

[0079] Example 1

[0080] An embodiment of the present application provides a beauty instrument control method, which is configured to ensure the safety of using the beauty instrument based on the relative position of the beauty instrument and the skin when the user uses the beauty instrument, thereby reducing the risk of the beauty instrument scalding the user's skin.

[0081] Please refer to FIG1 . The beauty instrument control method may include steps S101 to S104 , and each step is described in detail below.

[0082] Step S101: using the touch chip of the beauty instrument to identify whether the end face of the beauty instrument is in contact with the skin.

[0083] In this embodiment, a touch chip is installed within the front end of the beauty device. The voltage level signals collected by the touch chip can be used to determine whether the front end of the beauty device is in contact with the user's skin. When the voltage level signal collected by the touch chip is a high level signal, it indicates that the end face of the beauty device is in contact with the user's skin; when the voltage level signal collected by the touch chip is a low level signal, it indicates that the end face of the beauty device is not in contact with the user's skin. This method of using the touch chip to determine whether the beauty device is in contact with the skin is more accurate. In other embodiments, pressure sensors, optical sensors, thermal sensors, and other methods can also be selected according to actual needs to determine whether the beauty device is in contact with the skin.

[0084] Step S102 : If the end face of the beauty instrument is in contact with the skin, the beauty instrument load is started, and the camera of the beauty instrument is used to detect whether relative motion occurs between the skin and the end face of the beauty instrument based on motion estimation technology.

[0085] In this embodiment, if the beauty device's end face is detected in contact with the user's skin, the device's load is activated, restoring its rated voltage and entering normal operation. Furthermore, the device uses images captured by a camera mounted on the device and motion estimation technology to detect relative motion between the user's skin and the device's end face. This allows the device's operating state to be controlled based on the presence of relative motion between the device's end face and the skin, thereby reducing the risk of skin burns. The positional relationship between the camera and the device is shown in Figure 2, with camera 201 mounted on the inner side of the device's end face 202.

[0086] It should be noted that the working principle of a beauty instrument is to emit an infrared spectrum that is configured to penetrate deep into the skin. After the infrared is emitted, it will heat up when it encounters the receiving skin. At this time, there is a certain amount of heat. If the current movement state between the beauty instrument and the skin cannot be determined, and the beauty instrument always outputs energy based on full power, there will be a risk of scalding the skin. If the current movement state between the beauty instrument and the skin, or even the movement speed of the beauty instrument, can be accurately judged, the size of the infrared energy output of the beauty instrument can be adjusted in real time, thereby reducing the risk of skin burns.

[0087] In a specific embodiment, after identifying whether the beauty instrument is in contact with the skin through the touch chip in the beauty instrument, the method further includes: if the beauty instrument is not in contact with the skin, re-performing the step of identifying whether the end face of the beauty instrument is in contact with the skin through the touch chip in the beauty instrument.

[0088] In this embodiment, if the touch chip in the beauty device detects whether the beauty device is in contact with the skin and the result is that the beauty device is not in contact with the skin, the touch chip will continue to detect whether the beauty device is in contact with the skin until it is determined that the beauty device is in contact with the skin. Only when the beauty device is in contact with the skin can the detection of the relative motion state be performed.

[0089] In a specific embodiment, referring to FIG3 , the detection of whether the skin and the end face of the beauty instrument are in relative motion by the camera of the beauty instrument based on motion estimation technology includes steps S1021 to S1024 , each of which is described in detail below.

[0090] Step S1021: obtaining, through the camera, an image of the state of the contact surface between the skin and the end face of the beauty instrument.

[0091] In this embodiment, after recognizing that the end face of the beauty instrument is in contact with the skin, a camera is used to obtain an image of the contact surface between the skin and the end face of the beauty instrument, wherein the contact surface is the surface where the end face of the beauty instrument is in contact with the skin.

[0092] Step S1022: Determine whether there is pixel offset in the state image based on the motion estimation technology.

[0093] In this embodiment, motion estimation technology is used to determine whether pixel offsets exist in the state image, thereby improving image tracking accuracy. The motion estimation technology can be an optical flow slide detection algorithm, which can be further divided into dense and sparse methods. These algorithms are widely used in motion analysis, target tracking, video compression, and other fields.

[0094] In a specific embodiment, the determining whether there is pixel offset in the state image based on the motion estimation technology includes: calculating the displacement vector corresponding to each pixel point in the state image of the current frame and each pixel point in the state image of the previous frame through the motion estimation technology; determining whether the displacement vector corresponding to each pixel point is greater than a displacement threshold; if there is a displacement vector greater than the displacement threshold, then there is pixel offset in the state image; if there is no displacement vector greater than the displacement threshold, then there is no pixel offset in the state image.

[0095] In this embodiment, motion estimation technology is used to calculate the displacement vectors corresponding to each pixel in the state image of the current frame and each pixel in the state image of the previous frame. A determination is made as to whether the displacement vector corresponding to each pixel is greater than a displacement threshold. If a displacement vector is greater than the displacement threshold, pixel shift exists in the state image. If a displacement vector is not greater than the displacement threshold, pixel shift does not exist in the state image. The displacement threshold is set based on actual needs and is not limited in this embodiment.

[0096] Step S1023: If there is pixel offset in the status image, the skin and the end surface of the beauty instrument undergo relative motion.

[0097] In this embodiment, if there is a pixel shift between the state image of the current frame and the state image of the previous frame, it can be known that relative movement has occurred between the skin and the end face of the beauty instrument.

[0098] Step S1024: If there is no pixel offset in the status image, then there is no relative movement between the skin and the end surface of the beauty instrument.

[0099] In this embodiment, if the state image of the current frame is compared with the state image of the previous frame, the skin and the end surface of the beauty instrument do not move relative to each other.

[0100] Step S103: If there is no relative movement between the skin and the end face of the beauty instrument, the power of the beauty instrument load is reduced until the beauty instrument is in a standby state; and the step of identifying whether the end face of the beauty instrument is in contact with the skin by the touch chip of the beauty instrument is re-executed.

[0101] In this embodiment, when it detects that there is no relative motion between the skin and the end face of the beauty device, the power of the beauty device load is reduced until the beauty device enters a standby state. The process then proceeds to step S101 to re-identify whether the end face of the beauty device is in contact with the skin. At this point, the beauty device stops continuously delivering energy at full power to the same part of the user's skin, thereby reducing the risk of skin burns. The device then continues to determine the state of the end face of the beauty device and the skin, allowing the beauty device to determine whether to continue operation based on this subsequent state, thereby improving the user experience.

[0102] Step S104: If the skin and the end surface of the beauty instrument move relative to each other, the beauty instrument is controlled to operate normally.

[0103] In this embodiment, when relative movement is detected between the skin and the end face of the beauty instrument, the beauty instrument is controlled to work normally. At this time, the beauty instrument continues to output energy, so that the beauty instrument releases energy to heat another part of the user's skin, thereby reducing the risk of skin burns caused by heating the same part of the skin.

[0104] In a specific embodiment, controlling the beauty instrument to operate normally includes: detecting the relative movement speed of the end face of the beauty instrument based on the motion estimation technology; outputting a target level signal according to the relative movement speed; adjusting the power of the beauty instrument load according to the target level signal; and controlling the beauty instrument to operate according to the power of the beauty instrument load.

[0105] In this embodiment, when relative motion between the skin and the end face of the beauty instrument is detected, the relative motion speed of the end face of the beauty instrument on the user's skin is continuously detected using motion estimation technology.

[0106] Specifically, the first system determines the relative movement speed of the end face of the beauty instrument based on the status image obtained by the camera through motion estimation technology, converts the analysis result of the relative movement speed into a target level signal, and sends it to the second system; the second system controls the power of the beauty instrument load based on the received target level signal, wherein the relative movement speed value determines the value of the output target level signal, and then determines the change in the power of the beauty instrument load according to the value of the target level signal.

[0107] For example, using 10 mm / s as a threshold, if the relative speed is above 10 mm / s, the rated power output in normal mode is used. If the relative speed is 10 mm / s, the target level signal value sent by the first system to the second system is coded as 5. Each time the speed decreases by N 2 mm / s, the value is coded as 5-N, until it reaches 0. The power is then progressively reduced according to the corresponding code. The specific mapping relationship between the relative speed value, the target level signal value, and the load power change can be set according to actual needs and is not limited in this embodiment.

[0108] In a specific embodiment, after the beauty instrument is controlled to operate normally, the control includes: detecting the end surface temperature of the end surface of the beauty instrument and determining the temperature change of the end surface temperature; if the temperature change is an increase, increasing the cooling power of the beauty instrument through a control algorithm; if the temperature change is unchanged, controlling the cooling power to remain unchanged; if the temperature change is a decrease, reducing the cooling power of the beauty instrument through a control algorithm.

[0109] In this embodiment, during normal operation of the beauty device, the device's end surface temperature is detected and its current temperature change is determined. If the end surface temperature increases, the control algorithm increases the power of the device's cooling chip, thereby reducing the end surface temperature to an appropriate level. If the end surface temperature decreases, the control algorithm decreases the power of the cooling chip, thereby increasing the end surface temperature to an appropriate level. This approach, by adjusting the cooling chip power to maintain the appropriate end surface temperature, ensures that the skin maintains a constant temperature and completes skin care.

[0110] In a specific embodiment, after increasing the power of the refrigeration plate of the beauty instrument through the control algorithm, it includes: reducing the power of the beauty instrument load; after reducing the power of the refrigeration plate of the beauty instrument through the control algorithm, it includes: increasing the power of the beauty instrument load.

[0111] In this embodiment, if adjusting the cooling chip power fails to adjust the end surface temperature of the beauty instrument to the appropriate range, it is necessary to adjust the end surface temperature by changing the power of the beauty instrument load. Therefore, after increasing the cooling chip power of the beauty instrument through the control algorithm, the power of the beauty instrument load can also be reduced, thereby reducing the end surface temperature to the appropriate range; after reducing the cooling chip power of the beauty instrument through the control algorithm, the power of the beauty instrument load can also be increased, thereby raising the end surface temperature to the appropriate range, thereby ensuring that the user performs skin care work at the appropriate temperature.

[0112] The beauty device control method proposed in this embodiment uses the beauty device's touch chip to identify whether the beauty device's end face is in contact with the skin. If so, the beauty device load is activated, and the beauty device's camera, using motion estimation technology, detects whether the skin and the beauty device's end face are in relative motion. If no relative motion occurs between the skin and the beauty device's end face, the power of the beauty device load is reduced until the beauty device enters a standby state. The process of identifying whether the beauty device's end face is in contact with the skin using the beauty device's touch chip is then repeated. If relative motion occurs between the skin and the beauty device's end face, the beauty device is controlled to operate normally. In this way, the energy release of the beauty device is controlled based on the touch chip's recognition results and the presence of relative motion between the beauty device's end face and the user's skin, thereby greatly improving the safety of the beauty device's energy output and reducing the risk of burns to the user's skin.

[0113] In addition, an embodiment of the present disclosure provides a beauty instrument control device, as shown in FIG4 , which includes:

[0114] Identification module 401 is configured to identify whether the end face of the beauty instrument is in contact with the skin through the touch chip of the beauty instrument;

[0115] The detection module 402 is configured to activate the beauty device load if the beauty device end face is in contact with the skin, and detect whether the skin and the beauty device end face are in relative motion using a camera based on motion estimation technology;

[0116] The first control module 403 is configured to reduce the power of the load until the beauty instrument is in a standby state if there is no relative movement between the skin and the end face of the beauty instrument; and re-execute the step of identifying whether the end face of the beauty instrument is in contact with the skin through the touch chip of the beauty instrument;

[0117] The second control module 404 is configured to control the beauty instrument to operate normally if the skin and the end surface of the beauty instrument move relative to each other.

[0118] Optionally, the detection module 402 is further configured to obtain a status image of the fitting surface between the skin and the end face of the beauty instrument through the camera; determine whether there is pixel offset in the status image based on the motion estimation technology; if there is pixel offset in the status image, then the skin and the end face of the beauty instrument undergo relative motion; if there is no pixel offset in the status image, then the skin and the end face of the beauty instrument do not undergo relative motion.

[0119] Optionally, the detection module 402 is further configured to calculate the displacement vector corresponding to each pixel point in the state image of the current frame and each pixel point in the state image of the previous frame through the motion estimation technology; determine whether the displacement vector corresponding to each pixel point is greater than the displacement threshold; if there is a displacement vector greater than the displacement threshold, then there is a pixel offset in the state image; if there is no displacement vector greater than the displacement threshold, then there is no pixel offset in the state image.

[0120] Optionally, the identification module 401 is further configured to re-execute the step of identifying whether the end face of the beauty instrument is in contact with the skin through the touch chip of the beauty instrument if the beauty instrument is not in contact with the skin.

[0121] Optionally, the second control module 404 is further configured to detect the relative movement speed of the end face of the beauty instrument based on the motion estimation technology; output a target level signal according to the relative movement speed; adjust the power of the beauty instrument load according to the target level signal; and control the beauty instrument to operate according to the power of the beauty instrument load.

[0122] Optionally, the second control module 404 is further configured to detect the end surface temperature of the beauty instrument and determine the temperature change of the end surface temperature; if the temperature change is an increase, the cooling power of the beauty instrument is increased through the control algorithm; if the temperature change is unchanged, the cooling power is controlled to remain unchanged; if the temperature change is a decrease, the cooling power of the beauty instrument is reduced through the control algorithm.

[0123] Optionally, the second control module 404 is further configured to reduce the power of the beauty instrument load after increasing the power of the refrigeration plate of the beauty instrument through a control algorithm;

[0124] The second control module 404 is further configured to increase the power of the beauty instrument load after reducing the power of the cooling plate of the beauty instrument through a control algorithm.

[0125] The device provided in the embodiment of the present disclosure can execute the steps of the beauty instrument control method provided above, which will not be described again to avoid repetition.

[0126] Example 2

[0127] The present invention provides a method for processing skin data. The method is configured as a skin beauty device with a high-speed multispectral imaging module at the front end. Referring to FIG5 , a flow chart of a method for processing skin data is shown. The method mainly includes the following steps S502 to S506:

[0128] Step S502: Acquire the current detailed skin image captured by the high-speed multispectral imaging module.

[0129] The high-speed multispectral imaging module can be located at the front end of the skin beauty device. When the user holds the skin beauty device and moves it across the human face, the high-speed multispectral imaging module can perform high-speed multispectral CMOS (Complementary Metal Oxide Semiconductor) image acquisition of the current skin area. The multispectral imaging picture acquired in real time is the above-mentioned current detailed skin image.

[0130] Step S504 : determining the current skin position of the skin beauty device by comparing the current detailed skin image with the pre-collected full-face skin image.

[0131] The full-face skin image can be acquired using a binocular multispectral imaging module, which can also be located on the front end of the skin beauty device. Before using the skin beauty device for skin care, the multispectral imaging module can be turned on, and then the skin beauty device can be held to scan the face. During the scanning process, the binocular CMOS multispectral imaging of the multispectral imaging module can be used to obtain a full-face image sample, i.e., a full-face skin image. The skin beauty device can then be used for skin care.

[0132] In some possible embodiments, the above-mentioned step S504 can be implemented through the following process: first obtain the current facial pattern code corresponding to the current detailed skin image; then match the current facial pattern code with multiple original facial pattern codes corresponding to the full-face skin image; and then determine the skin position corresponding to the successfully matched original facial pattern code as the current skin position of the skin beauty instrument.

[0133] Optionally, the current facial print code includes a feature vector. Based on this, the current facial print code can be obtained through the following process: first, current skin feature data in the current detailed skin image is obtained; wherein the current skin feature data includes attribute data under preset parameters, the preset parameters include one or more of pores, spots, and texture, and the attributes corresponding to the attribute data include one or more of color, size, position, distance, and area; then, based on the current skin feature data, a feature vector corresponding to the current detailed skin image is generated. Furthermore, when generating the feature vector corresponding to the current detailed skin image, the current skin feature data can be first quantized, and then the quantized data can be hexadecimalized, that is, converted into computer-readable data to obtain the feature vector.

[0134] In order to further improve the matching accuracy, facial print codes can be matched in combination with skin impedance values. Skin impedance values ​​can be collected by an electrode device installed at the front end of the skin beauty instrument. Each original facial print code corresponds to a skin impedance value range. The skin impedance value range is determined by the skin impedance value detected in the skin area where the original facial print code is located. For example, the maximum and minimum skin impedance values ​​in the skin area where the original facial print code is located are determined, and the range formed by the maximum and minimum values ​​is used as the skin impedance value range corresponding to the original facial print code. The size of human skin impedance is directly related to the thickness of the epidermis and dermis. If the skin impedance is small, it can be determined that the skin of the face is thin, and vice versa. Therefore, the skin impedance value can be used as a prerequisite for partition screening to scan the original facial print code corresponding to the current detailed skin image. Based on this, the specific matching process can be as follows: first, obtain the current skin impedance value collected by the electrode device; traverse each original facial pattern code; for the traversed original facial pattern code, if the current skin impedance value is within the preset interval of the original facial pattern code and the original facial pattern code can be identified, the match is successful and the matching ends; if the current skin impedance value is not within the preset interval of the original facial pattern code, but can match the original facial pattern code, it is regarded as an incorrect selection or interference, the match fails, and the matching of the next original facial pattern code is continued; if the current skin impedance value and the original facial pattern code cannot be paired, the match fails, and the matching of the next original facial pattern code is continued; until the matching ends.

[0135] Step S506: determining current treatment parameters of the skin beauty device according to the current detailed skin image and the current skin position.

[0136] The current treatment parameters may include one or more of the amount of skin care solution used, the permeability of the skin care solution, the frequency of the activation signal, and the energy of the activation signal. The activation signal may include one or more of a radio frequency signal, a low frequency signal, and red light. The skin care solution may be a liquid skin care product or a liquid medicine with skin therapeutic effects. The specific setting can be based on actual needs and is not limited here.

[0137] The embodiments of the present application provide three implementation methods for determining the current treatment parameters of a skin beauty device. These three implementation methods are introduced below.

[0138] Implementation method 1:

[0139] First, the system obtains the previous detailed skin image and previous treatment parameters corresponding to the current skin location. The previous detailed skin image is the detailed skin image captured by the high-speed multispectral imaging module at the current skin location before the most recent treatment, and the previous treatment parameters are the treatment parameters corresponding to the current skin location during the most recent treatment. The system then determines the previous treatment effect based on the current and previous detailed skin images. Finally, the current treatment parameters of the skin beauty device are determined based on the previous treatment effect and parameters. This approach references the user's previous treatment results, resulting in more accurate current treatment parameters.

[0140] Optionally, the previous treatment effect can be determined by the following process: obtaining the current skin feature data in the current detailed skin image and the previous skin feature data in the previous detailed skin image; comparing the current skin feature data with the previous skin feature data to obtain a comparison result; and determining the previous treatment effect based on the comparison result. Furthermore, when determining the previous treatment effect based on the comparison result, the change amount of each preset parameter can be obtained first (the change amount can be a change ratio, including three cases of positive, 0, and negative numbers), and then the change amount of each preset parameter is weighted and summed to obtain an effect evaluation value, and then the previous treatment effect is determined based on the effect evaluation value. For example, when the effect evaluation value is positive, the previous treatment effect is a positive change; when the effect evaluation value is 0, the previous treatment effect is no change; when the effect evaluation value is negative, the previous treatment effect is a negative change.

[0141] Optionally, the last treatment effect includes a positive change, no change, or a negative change; based on this, the step of determining the current treatment parameter of the skin beauty instrument according to the last treatment effect and the last treatment parameter can be performed as follows: when the last treatment effect is a positive change, the last treatment parameter is determined as the current treatment parameter of the skin beauty instrument; when the last treatment effect is no change, the treatment parameter obtained by increasing the last treatment parameter by a first preset ratio is determined as the current treatment parameter of the skin beauty instrument; when the last treatment effect is a negative change, the treatment parameter obtained by reducing the last treatment parameter by a second preset ratio is determined as the current treatment parameter of the skin beauty instrument. Among them, the first preset ratio and the second preset ratio can be set according to actual needs and are not limited here.

[0142] Implementation method 2:

[0143] First, the preset basic treatment parameters and current skin impedance value corresponding to the current skin location are obtained; then, the preset basic treatment parameters are adjusted according to the current skin impedance value to obtain the current treatment parameters of the skin beauty device. The preset basic treatment parameters can be treatment parameters statistically obtained based on a large amount of experimental data, and the preset basic treatment parameters corresponding to different parts of the body are usually different. This method does not require reference to the user's previous treatment results and has a certain degree of versatility. It is particularly suitable for scenarios where a skin beauty device is used for skin care for the first time.

[0144] Furthermore, to improve the accuracy of the current treatment parameters, the current skin impedance value can be combined with the current skin feature data in the current detailed skin image to adjust the preset basic treatment parameters. Alternatively, the preset basic treatment parameters can be adjusted directly based on the current skin feature data in the current detailed skin image. The specific adjustment strategy can be set according to actual needs and is not limited here.

[0145] Implementation method three:

[0146] The current detailed skin image and the current skin position are input into a data model obtained based on deep learning, and the treatment parameters output by the data model are used as the current treatment parameters of the skin beauty device; wherein the data model is trained based on historical detailed skin images and historical treatment parameters corresponding to different skin positions of the current user.

[0147] The entire data model training process includes four parts: dataset construction, model training, model loading, and model parameter adjustment. The dataset includes a training set and a validation set, each of which includes multiple samples. Each sample includes historical detailed skin images and historical treatment parameters corresponding to different skin locations of the user. The historical detailed skin images serve as input to the data model, and the historical treatment parameters serve as labels. The data model training process can be as follows: training on the training set; validation on the validation set; saving the optimal weights for the data model and reading the weights; and recording the accuracy of the training and validation sets to facilitate parameter adjustment.

[0148] The skin data processing method provided in the embodiment of the present application adopts high-speed multispectral imaging acquisition technology combined with a pre-collected full-face skin image to quickly complete the recognition and positioning navigation functions of the user's facial skin details; based on the current detailed skin image and the current skin position, the current treatment parameters of the skin beauty instrument are adjusted so that different skin positions and different skin conditions correspond to different treatment parameters, thereby achieving precise skin care.

[0149] Furthermore, the current detailed skin image, current skin position and current treatment parameters can also be sent to a preset processing device so that the processing device generates a skin quantitative data report; wherein, the processing device can be the user terminal and / or cloud server corresponding to the skin beauty instrument.

[0150] The skin quantitative indicators in the above-mentioned skin quantitative data report may include: pore color, spot area, number of texture lines and capillary depth of the skin captured by the multispectral imaging module. The user terminal can be an APP, and the user can view the skin quantitative data report in the APP. Taking the user terminal as an APP as an example, when the user opens the user APP, based on the various skin positioning points (i.e. skin positions or skin points) obtained by the front end of the skin beauty instrument and the reference quantity of the full face image analysis, the status of each skin point is determined (the status may include inflammation, spot depth, distribution of capillaries, sensitive skin red areas, moisture, oil, etc.), and detailed skin images of each point are obtained and transmitted to the user APP for local processing to form a skin quantitative data report; when the user does not open the user APP, it can be directly transmitted to the cloud server for processing, and the skin quantitative data report is sent back to the local background of the user APP when the user APP is started.

[0151] For ease of understanding, the above-mentioned skin data processing method is described in detail below with reference to FIG. 6 .

[0152] As shown in Figure 6, the front end of the skin beauty instrument is equipped with a high-speed multispectral imaging module and a binocular multispectral imaging module. Among them, the high-speed multispectral imaging module obtains multispectral imaging pictures through high-speed multispectral CMOS image acquisition. The high-speed multispectral imaging module is configured to collect skin detail points and analyze skin subtle imaging; the binocular multispectral imaging module obtains full-face image samples through binocular CMOS multispectral imaging. The binocular multispectral imaging module is configured to collect full-face image positioning points; the multispectral imaging pictures and full-face image samples are input into the CPU (Central Processing Unit) / MCU (Microcontroller Unit), and the user's full-face image is used as a reference to establish an image archive of the full-face skin. These image files are generated into facial pattern codes and stored.

[0153] The CPU / MCU can transmit the acquired image to the APP or cloud server, and can transmit the image through hardware encoding and decoding, and process it locally on the APP or on the cloud server to generate a skin quantitative data report. Specifically, the CPU / MCU can determine the status of each skin point (including inflammation, spot depth, distribution of capillaries, red areas of sensitive skin, moisture, oil, etc.) based on the obtained skin points and the reference amount of full-face image analysis, obtain detailed skin images of each point, and transmit them to the user's APP for local processing to form a skin quantitative data report. The skin quantitative indicators in the skin quantitative data report include: pore color, spot area, number of texture lines, and capillary depth of the skin captured by the multispectral imaging module; or when the user does not open the APP, it is directly transmitted to the cloud server for processing, and the data is sent back to the user's local APP background when the APP is started.

[0154] Using the full-face image as a reference surface and the multispectral imaging image as a positioning point, facial skin navigation is performed to determine the skin point where the skin beauty device is located. The facial skin navigation process can be as follows: using the image captured by the high-speed multispectral imaging module as a reference point, a detailed skin image of the current face is acquired. The facial print code of the current detailed skin image is then retrieved and compared with the facial print codes of the full-face images in the archive. The sampled detailed skin image is then compared with the stored full-face images in sequence to find the best match. The skin location corresponding to this match is the navigation result. The facial print code can be obtained by first determining the color, size, position, distance, and area of ​​pores, spots, and textures in the facial image (i.e., the detailed skin image), and then calculating their geometric features. These features form a feature vector describing the facial image. The facial print code of the detailed skin image is used as an identification parameter to compare, judge, and confirm with all the original parameters in the database (i.e., the archive). Ultimately, the current skin location of the skin beauty device front end is determined, completing facial skin navigation.

[0155] After completing the sampling of each skin point and facial skin navigation, the skin beauty instrument begins to be adjusted. The adjustments include activation signals such as RF (radio frequency), EMS (microcurrent or low frequency) and red light, and their energy and frequency are adjusted so as to be configured to improve reasonable skin conditions. That is, according to the navigation results, the energy and frequency output are adjusted to configure the skin. By adjusting the energy and frequency output of the activation signal, the penetration and absorption rate of skin care ingredients at different skin points can be changed. Also, according to the navigation results, the dosage and administration time of each skin point are changed (the dosage is the amount of skin care solution when it is a liquid medicine). By changing the dosage and administration time of each skin point, combined with the adjusted energy and frequency output of the activation signal, the absorption efficiency of each skin point can be adjusted to achieve a balanced skin effect and avoid uneven skin conditions.

[0156] This embodiment of the application provides two adjustment methods, namely:

[0157] Method 1: Based on the parameters corresponding to the previous treatment (parameters include the amount of skin care solution used, the penetration rate of the skin care solution, the energy of the activation signal, and the frequency of the activation signal) and the treatment effect (the treatment effect can be determined based on the skin condition before and after the previous treatment), the parameters corresponding to the current treatment are determined. The specific implementation method can be as follows:

[0158] Based on the previous skin data attributes, including the color, size, position, distance, area and other attributes of the facial skin such as pores, spots, texture, etc. after multi-spectral irradiation, the data of the quantified skin characteristics are compared with the skin data obtained this time to obtain the change ratio; there are three situations: positive change, no change and negative change. For positive changes, the treatment parameters are kept unchanged from the previous treatment; for no change, the energy and frequency, the amount of medicine and the penetration rate of the liquid medicine are increased by 10% again; for negative changes, the frequency and energy are reduced by 10%, and the amount of medicine is reduced, and the user is prompted with possible risks.

[0159] Method 2: Based on the sampled detailed skin image and real-time skin impedance value, the corresponding parameters of this treatment are determined. The specific implementation method can be as follows:

[0160] When collecting skin data for the first time, since there is no previous reference for comparison, method 1 cannot be used to obtain the adjustment conditions. Therefore, the energy and frequency outputs are adjusted based on the skin feature data and real-time skin impedance corresponding to the detailed skin image. The default reference amount for drugs remains unchanged.

[0161] The above-mentioned dosage, absorption efficiency, and changes in the energy and frequency of RF, EMS and red light can also be transmitted to the APP for local processing to further improve the user's skin data report and continuously improve the large database.

[0162] Furthermore, after completing facial skin navigation, real-time image data of skin changes and drug dosage at various points on the skin are obtained, and a data model is established through enhanced deep learning. This allows for continuous improvement of the database, establishment of data models, and deep learning to complete the collection and update of skincare clinical data. Among these, skin pores, spots, texture, and capillaries, corresponding to the color, size, location, distance, and area of ​​the entire face, have been analyzed. The rate of change of corresponding skin attribute characteristic data indicators under the influence and care of RF, EMS, red light, and drugs, as well as within a certain treatment or care cycle, forms the data model.

[0163] Through the above data model, deep learning training is strengthened to provide skin beauty devices with faster response to match the user's skin, output the corresponding energy, frequency and matching skin care solution and dosage. At the same time, different skin care solution ingredients can be classified and counted according to the skin conditions of different users of different genders and ages, further improving the formula of skin care solution.

[0164] This application has built a database of thousands of images of skin conditions, laying a solid foundation for deep learning to identify skin conditions and establish drug clinical data and drug tracking, making skin recognition and visualization more accurate and less error-tolerant.

[0165] This application uses high-speed multispectral imaging acquisition technology, which can quickly confirm the pixel positioning points on the user's entire face, play a navigation role, and realize real-time positioning and skin detection. The command operation process of facial image navigation and positioning combines the multispectral binocular large camera imaging acquisition of the device body and the high-speed small camera facial detail recognition to complete the recognition and positioning navigation function of the user's facial skin details; the status of each point of the skin is collected in a specified manner, which can accurately judge the condition of the skin, including the environment inside the skin's pores, the condition of capillaries that are invisible to the naked eye, deep skin lesions, etc., thereby providing a reference for the various parameter outputs of the device in real time, making different adjustments to the data output according to different skin conditions, so that the skin can obtain the effects of precise skin care and precise drug delivery, and achieve balanced skin care;

[0166] Furthermore, the acquired data will be used to continuously train data models, strengthen deep learning, continuously upgrade and reduce fault tolerance, so that the equipment is always in the latest state;

[0167] Furthermore, we will establish a database for quantifying skin indicators, combine it with users' clinical care reports, optimize drug formulations and dosages, and use the acquired data to establish classification statistics for different drug ingredients corresponding to the skin conditions of different users' genders and ages, further improve drug formulations, and provide practical and effective clinical care data.

[0168] Corresponding to the above-mentioned skin data processing method, the present application also provides a skin data processing device. The device should also be configured as a skin beauty device with a high-speed multispectral imaging module at the front end. Referring to the structural diagram of a skin data processing device shown in FIG7 , the device includes:

[0169] An image acquisition module 701 is configured to acquire a current detailed skin image acquired by a high-speed multispectral imaging module;

[0170] a position determination module 702 configured to determine the current skin position of the skin beauty device by comparing the current detailed skin image with a pre-collected full-face skin image;

[0171] The parameter determination module 703 is configured to determine the current treatment parameters of the skin beauty device according to the current detailed skin image and the current skin position.

[0172] Furthermore, the above-mentioned position determination module 702 is specifically configured to: obtain the current facial pattern code corresponding to the current detailed skin image; match the current facial pattern code with multiple original facial pattern codes corresponding to the full-face skin image; and determine the skin position corresponding to the successfully matched original facial pattern code as the current skin position of the skin beauty instrument.

[0173] Furthermore, the above-mentioned current facial print encoding includes a feature vector; based on this, the above-mentioned position determination module 702 is also configured to: obtain current skin feature data in the current detailed skin image; wherein, the current skin feature data includes attribute data under preset parameters, the preset parameters include one or more of pores, spots and textures, and the attributes corresponding to the attribute data include one or more of color, size, position, distance and area; according to the current skin feature data, generate a feature vector corresponding to the current detailed skin image.

[0174] In some possible embodiments, the parameter determination module 703 is specifically configured to: obtain the previous detailed skin image and the previous treatment parameters corresponding to the current skin position; wherein the previous detailed skin image is the detailed skin image captured by the high-speed multispectral imaging module at the current skin position before the most recent treatment, and the previous treatment parameters are the treatment parameters corresponding to the current skin position during the most recent treatment; determine the previous treatment effect based on the current detailed skin image and the previous detailed skin image; determine the current treatment parameters of the skin beauty instrument based on the previous treatment effect and the previous treatment parameters.

[0175] Furthermore, the parameter determination module 703 is also configured to: obtain the current skin feature data in the current detailed skin image and the previous skin feature data in the previous detailed skin image; compare the current skin feature data with the previous skin feature data to obtain a comparison result; and determine the previous treatment effect based on the comparison result.

[0176] Furthermore, the above-mentioned previous treatment effect includes positive change, no change or negative change; based on this, the above-mentioned parameter determination module 703 is also configured to: when the previous treatment effect is a positive change, the previous treatment parameter is determined as the current treatment parameter of the skin beauty instrument; when the previous treatment effect is no change, the treatment parameter obtained by increasing the previous treatment parameter by a first preset ratio is determined as the current treatment parameter of the skin beauty instrument; when the previous treatment effect is a negative change, the treatment parameter obtained by reducing the previous treatment parameter by a second preset ratio is determined as the current treatment parameter of the skin beauty instrument.

[0177] In other possible embodiments, the parameter determination module 703 is specifically configured to: obtain the preset basic treatment parameters and the current skin impedance value corresponding to the current skin position; adjust the preset basic treatment parameters according to the current skin impedance value to obtain the current treatment parameters of the skin beauty instrument.

[0178] In some other possible embodiments, the above-mentioned parameter determination module 703 is specifically configured to: input the current detailed skin image and the current skin position into a data model obtained based on deep learning, and use the treatment parameters output by the data model as the current treatment parameters of the skin beauty device; wherein the data model is trained based on historical detailed skin images and historical treatment parameters corresponding to different skin positions of the current user.

[0179] Furthermore, the skin data processing device further includes:

[0180] The data sending module is configured to send the current detailed skin image, current skin position and current treatment parameters to a preset processing device so that the processing device generates a skin quantitative data report; wherein the processing device is the user terminal and / or cloud server corresponding to the skin beauty instrument.

[0181] Furthermore, the above-mentioned full-face skin image is acquired through a binocular multispectral imaging module; the current treatment parameters include the amount of skin care solution, the penetration rate of the skin care solution, the frequency of the activation signal and the energy of the activation signal, and the activation signal includes one or more of radio frequency signals, low frequency and red light.

[0182] The skin data processing device provided in this embodiment has the same implementation principle and technical effects as those of the aforementioned skin data processing method embodiment. For the sake of brief description, for matters not mentioned in the skin data processing device embodiment, reference may be made to the corresponding content in the aforementioned skin data processing method embodiment.

[0183] As shown in Figure 8, an embodiment of the present application also provides a skin beauty instrument 800, including: a processor 801, a memory 802 and a bus. The memory 802 stores a computer program that can be run on the processor 801. When the skin beauty instrument 800 is running, the processor 801 and the memory 802 communicate through the bus, and the processor 801 executes the computer program to implement the above-mentioned skin data processing method.

[0184] Specifically, the memory 802 and processor 801 can be general-purpose memories and processors, which are not specifically limited here.

[0185] As shown in FIG9 , an embodiment of the present application further provides a skin beauty system, which includes the above-mentioned skin beauty instrument 901 , and also includes a user terminal 902 and a cloud server 903 respectively connected to the skin beauty instrument 901 .

[0186] As shown in Figure 9, the user terminal 902 can also be connected to the cloud server 903. The user terminal 902 can be understood as the user APP corresponding to the skin beauty instrument 901. The skin beauty instrument 901 can send detailed skin images of each detected point, corresponding skin positions and corresponding treatment parameters to the user APP. The user APP can process the data locally to form a skin quantitative data report; or when the user does not open the APP, the above data can be directly transmitted to the cloud server 903, and the cloud server 903 processes and generates a skin quantitative data report. When the APP is started, the data and / or the skin quantitative data report are sent back to the user's local APP background.

[0187] Example 3

[0188] This application provides a multi-dimensional control method for a beauty device, which should be configured as a server. Referring to FIG10 , a flowchart of a multi-dimensional control method for a beauty device is shown, which illustrates that the method mainly includes the following steps S1001 to S1004:

[0189] Step S1001: Acquire electrode signals, skin sampling data, and image data collected by the beauty device.

[0190] In an optional embodiment, the beauty device acquires electrode signals, skin sampling data and image data through electrodes, CMOS image sampling and imaging, and uploads the acquired electrode signals, skin sampling data and image data to a server.

[0191] Step S1002: Determine whether the beauty device is in contact with the skin of the current user based on the electrode signal.

[0192] In an optional embodiment, the server analyzes the collected electrode signals, converts them through an ADC, and then analyzes and determines whether the beauty device is in contact with the user's skin when the signal is high; and whether the signal is low. After obtaining a one-dimensional analysis result based on the electrode signals, the analysis result can be sent to the controller of the beauty device for further instructions.

[0193] Step S1003: If the skin of the current user is in contact, the working state of the beauty device is determined based on the skin sampling data.

[0194] In an optional embodiment, the beauty device samples the skin through a CMOS image, and the server determines whether the target pixel point has changed based on the target pixel point of the collected skin sampling data, thereby determining whether the beauty device is moving relative to the skin, and then converts the collected image into high and low level signals. If it is a low level signal, it indicates that the beauty device and the skin are in a relatively static state. If it is a high level signal, it is determined that the beauty device and the skin are in a state of relative motion. At the same time, the judgment is made in combination with the electrode signal. Under the condition that the electrode signal has contacted the skin, if the target pixel point has not changed, it is determined that the working state of the beauty device is in a fixed positioning stamping working mode. If the target pixel point changes, it is determined that the working state of the beauty device is in a sliding working mode.

[0195] Step S1004: Determine the skin condition of the current user based on the pre-trained skin detection model and image data.

[0196] In an optional embodiment, the server may input the image data into a pre-trained skin detection model to obtain the skin condition of the current user.

[0197] In specific implementations, the current user's skin condition can be determined using methods including, but not limited to, the following: inputting image data into a pre-trained skin detection model, annotating the skin type and skin type points in the image data, and obtaining the current user's skin condition; wherein the skin condition includes: the skin type and the points corresponding to the skin type, and the skin type includes one or more of the following: normal skin, acne, wrinkles, and spots. Specifically, the skin detection model can identify different skin types and points, and annotate the different skin types in the image data to obtain the skin condition.

[0198] The multi-dimensional control method of the above-mentioned beauty device provided in the embodiment of the present application judges and identifies the skin condition and the working status of the beauty device in multiple dimensions through electrode signals, skin sampling data and image data.

[0199] The present application also provides a method for training a skin detection model, as shown in FIG11 , which mainly includes the following steps 1 to 5:

[0200] Step 1: Acquire training images; wherein the training images include: marked skin images and unmarked skin images.

[0201] Step 2: Perform cross-domain image style transfer on training images.

[0202] In an optional embodiment, because differences in light sources and individual skin can lead to significant differences in image style, in the absence of labeled data, it is impossible to accurately detect new images using a model trained on a certain style image. Therefore, in the embodiments of this application, before model training, it is necessary to perform cross-domain image style transfer on the training images to reduce the distribution difference between the labeled domain (labeled skin images) and the unlabeled domain (unlabeled skin images).

[0203] Step 3: Train the prototype network based on the labeled skin images after cross-domain image style transfer to obtain the category prototype.

[0204] In an optional embodiment, a small number of labeled skin images in the labeled domain may be used to perform prototype network learning and establish category prototypes, that is, to classify the labeled skin images and determine different types of skin images.

[0205] Step 4: Use the cluster center nearest neighbor rule to match the unlabeled skin image with the category prototype to generate a pseudo label for the unlabeled skin image.

[0206] In an optional embodiment, the cluster center nearest neighbor rule can be used to match the unlabeled skin images with category prototypes, thereby clustering the unlabeled skin images and generating pseudo labels for the unlabeled skin images.

[0207] Step 5: Based on the pseudo labels of the unlabeled skin images, the skin detection model is trained to obtain a trained skin detection model.

[0208] In an optional embodiment, the skin detection model is trained using pseudo labels of unlabeled skin images to obtain a trained skin detection model, thereby improving the model's detection capability for images in the unlabeled domain. The skin detection model may be a skin detection model based on deep learning.

[0209] Furthermore, considering the differences between existing skin images and the skin images captured by the beauty device provided in the embodiments of this application, the skin detection model trained using existing skin images has lower accuracy when used to detect skin images captured by the beauty device provided in the embodiments of this application. Furthermore, the data volume of the skin images captured by the current beauty device is relatively small. Therefore, in the embodiments of this application, image reconstruction can be performed based on the skin images captured by the current beauty device to generate new training images. Specifically, the training data can be obtained using methods including but not limited to the following: first, obtaining an initial training image; then, inputting the initial training image into a feature extractor for feature extraction to obtain a feature image; and finally, inputting the feature image into an image generator for image reconstruction to obtain a training image.

[0210] As shown in FIG12 , the acquired initial training image is input into the feature extractor to obtain a feature image; the feature image is then input into the image generator to generate a new image to obtain a training image.

[0211] Specifically, the labeled dataset (i.e., labeled skin images) is The unlabeled dataset (i.e., unlabeled skin images) is N s and N t is the total number of samples in each data set, is the category label. From this, we can get the output of the feature extractor as:

[0212] Among them, θ f are the network parameters of the feature extractor. The specific structure of the feature extractor network is shown in Table 1.

[0213] In Table 1, the "7×7, 64, stride2" network has a 7×7 convolution kernel, 64 indicates that the number of output channels of this layer is 64, and stride 2 indicates that the number of pixels jumped when the convolution kernel slides is 2. The same applies to other network layers and will not be repeated here.

[0214] Table 1 Feature extractor Resnet-152 network

[0215] Specifically, the Transformer encoder network structure of the feature extractor in Table 1 can be seen in Figure 4, where the "Positional Encoding" operation is the position encoding operation:

[0216] Where, d model is the dimension of the input vector, which is 1000 in the embodiment of the present application, pos is the position of the image block, 2i+1 and 2i are odd and even values ​​from 1 to 1000.

[0217] In Figure 13, “Attention” is the attention mechanism operation: Q = X × W Q K=X×W K V=X×W V

[0218] Where W * represents the weight matrix, and X is the input vector. k is the vector dimension of K.

[0219] In the embodiment of the present application, the output of the feature extractor is used as the input of the image generator g, and the reconstructed image is and Then we have:

[0220] Where θ g are the learnable network parameters in the image generator.

[0221] From the above, we can see that the generator of the generative adversarial network G = f g. The generation loss function of the image generator is:

[0222] In the above formula, D is the discriminator in the generative adversarial network: a 101-layer residual network. The ideal discriminator's ability to distinguish images from unlabeled datasets is to judge them as 1 and images from labeled datasets as 0; the adversarial loss function is:

[0223] By L G and L D Through adversarial training, the style of the generated images will be biased towards the unlabeled dataset.

[0224] Furthermore, in order to optimize the feature extractor f and image generator g and retain the details of the unlabeled data, the following loss function is added:

[0225] In order to reduce the style difference of feature extractors, the feature extractor f and image generator g are further optimized. The features of the labeled dataset and the unlabeled dataset after passing through the feature extractor are:

[0226] The loss function of the feature extractor is the cross-domain loss function L CD :

[0227] Where, L MMD To maximize the mean difference loss function, L KL is the KL divergence loss function.

[0228] Through the above process, the feature extractor f and image generator g are continuously optimized, so that the style of the generated image will be biased towards the unlabeled dataset.

[0229] This embodiment of the present application also provides a process for generating pseudo labels for an unlabeled dataset, including:

[0230] (1) Training the prototype network using a labeled dataset:

[0231] Where μ C S is the category prototype, F′(xi θ f ) is the embedding function, and the prototype learning loss function is as follows:

[0232] (2) The prototype of the unlabeled dataset is expressed as:

[0233] (3) The pseudo labels of the unlabeled dataset are: Right now with c j The prototype is the best match.

[0234] Furthermore, the feature extractor f is further optimized according to the pseudo labels and true labels. First, prototype training is performed based on the pseudo labels to obtain the prototype of the unlabeled dataset:

[0235] Then, we further optimize f through the class-level MMD loss function to reduce the distribution differences between the generated datasets. The MMD loss function is as follows:

[0236] Furthermore, softmax is added to the feature extractor for classification prediction, and the corresponding classification loss function is:

[0237] The labeled data set used in the embodiments of the present application includes a positive data set and a negative data set. The positive data set can be the SFA public data set, and the negative data set can be extracted from the "Oriental Face Library"; the unlabeled data set can be images taken by the imaging system of the beauty equipment.

[0238] An embodiment of the present application provides another multi-dimensional control method for a beauty device. This method can be applied to beauty devices. The beauty device can also utilize a lightweight skin classification model deployed on the device to implement skin quality assessment and achieve dynamic control of the device output power. Specifically, the method includes: first, determining the skin condition of the current user based on a pre-trained skin classification model and image data; then, determining the output power of the beauty device based on the skin condition.

[0239] Referring to the schematic diagram of a lightweight skin classification model shown in FIG14 , it is illustrated that the skin classification model includes at least: an image block embedding module, multiple Vit sub-networks, multiple downsampling modules, an average pooling module and a classification module.

[0240] In an optional embodiment, when determining the skin condition of the current user based on a pre-trained skin classification model and image data, the following methods may be used, including but not limited to: inputting the image data into a pre-trained skin classification model, performing feature extraction and position extraction on the image data through an image block embedding module to obtain feature information and position information; classifying the feature information and position information in sequence through a Vit subnetwork, a downsampling module, an average pooling module, and a classification module to obtain the skin condition of the current user; wherein, the skin condition includes: skin type and points corresponding to the skin type, and the skin type includes one or more of the following: normal skin, acne, wrinkles, and spots.

[0241] In an alternative embodiment, the skin classification model primarily detects the following four skin types and their corresponding locations: 1) normal skin, 2) acne, 3) wrinkles, and 4) spots. To enable edge devices to accurately detect skin quality, a lightweight skin classification model, as shown in Figure 15, is used for detection.

[0242] As shown in Figure 14, the image block embedding module is configured to extract image block features and block position information. The Vit sub-network implements the attention mechanism. The specific structure of the Vit sub-network is shown in Figure 15.

[0243] The network expression of Vit sub-network is: X i+1 =Φ F (Φ A (Φ F′ (X i )))

[0244] Where, X i is the model input, X i+1 is the model output, Φ F′ represents the depth separation convolution and feedforward neural network at the input end, Φ A represents the self-attention subnetwork, Φ F Represents the deep separation convolution and feedforward neural network at the output end; the self-attention subnetwork structure is shown in Figure 16. The self-attention subnetwork first segments the input image, then processes each segmented image separately, and finally combines the features of the segmented images to obtain the final output result.

[0245] The self-attention subnetwork is expressed as follows: Y=Concat[Y ij ] j＝1∶h W P

[0246] Where, X ij is the feature that enters the jth cascade network after segmentation, Y ijis the output of the jth cascade network. “concat” is the feature combination operation, W p is the weight matrix of the fully connected layer network, and Y is the classification result.

[0247] The training dataset for the lightweight skin classification model provided in this embodiment includes positive and negative samples. The positive dataset consists of: 1) the SFA public dataset and 2) images captured by beauty devices; the negative dataset consists of: 1) images captured from the "Oriental Face Database" and 2) images captured by beauty devices.

[0248] The above method provided in this embodiment identifies the skin condition through a lightweight skin classification model and controls the beauty device according to the skin condition, thereby improving the accuracy of liquid discharge from the beauty device.

[0249] Regarding the aforementioned multi-dimensional control method for a beauty device, an embodiment of the present application provides a multi-dimensional control device for a beauty device, which is applied to a server. FIG. 17 shows a schematic diagram of the structure of a multi-dimensional control device for a beauty device, which mainly includes the following components:

[0250] The data acquisition module 1701 is configured to acquire electrode signals, skin sampling data, and image data collected by the beauty device;

[0251] an electrode signal determination module 1702 configured to determine whether the beauty device is in contact with the skin of the current user based on the electrode signal;

[0252] a working state determining module 1703 configured to determine the working state of the beauty device based on the skin sampling data if the skin of the current user is in contact;

[0253] The skin condition determination module 1704 is configured to determine the skin condition of the current user based on a pre-trained skin detection model and image data.

[0254] The multi-dimensional control device of the above-mentioned beauty device provided in the embodiment of the present application judges and identifies the skin condition and the working status of the beauty device in multiple dimensions through electrode signals, skin sampling data and image data.

[0255] Furthermore, the above-mentioned skin condition determination module 1704 is also configured to: input the image data into a pre-trained skin detection model, mark the skin type and skin type points in the image data, and obtain the skin condition of the current user; wherein the skin condition includes: skin type and skin type corresponding points, and the skin type includes one or more of the following: normal skin, acne, wrinkles and spots.

[0256] Furthermore, the above-mentioned device also includes a model training module, configured to: obtain training images; wherein the training images include: marked skin images and unmarked skin images; perform cross-domain image style transfer on the training images; perform prototype network training based on the marked skin images after cross-domain image style transfer to obtain category prototypes; use the cluster center nearest neighbor rule to match the unmarked skin images with the category prototypes to generate pseudo labels for the unmarked skin images; and train the skin detection model based on the pseudo labels of the unmarked skin images to obtain a trained skin detection model.

[0257] Furthermore, the above-mentioned device also includes a model training module configured to: obtain an initial training image; input the initial training image into a feature extractor for feature extraction to obtain a feature image; input the feature image into an image generator for image reconstruction to obtain a training image.

[0258] An embodiment of the present application provides another multi-dimensional control device for a beauty device, which is applied to the beauty device and includes: a power control module configured to: determine the skin condition of the current user based on a pre-trained skin classification model and image data; and determine the output power of the beauty device based on the skin condition.

[0259] Furthermore, the skin classification model includes at least: an image block embedding module, multiple Vit sub-networks, multiple downsampling modules, an average pooling module and a classification module.

[0260] Furthermore, the above-mentioned device also includes a power control module configured to: input image data into a pre-trained skin classification model, perform feature extraction and position extraction on the image data through the image block embedding module to obtain feature information and position information; classify the feature information and position information in sequence through the Vit sub-network, downsampling module, average pooling module and classification module to obtain the skin status of the current user; wherein, the skin status includes: skin type and the point corresponding to the skin type, and the skin type includes one or more of the following: normal skin, acne, wrinkles and spots.

[0261] The device provided in the embodiment of the present application has the same implementation principle and technical effects as those in the aforementioned method embodiment. For the sake of brief description, for matters not mentioned in the device embodiment, reference can be made to the corresponding content in the aforementioned method embodiment.

[0262] It should be noted that the specific numerical values ​​provided in the implementation of this application are only exemplary and are not intended to be limiting.

[0263] An embodiment of the present invention further provides a beauty device. Specifically, the electronic device includes a processor and a storage device; the storage device stores a computer program, and when the computer program is run by the processor, it executes the multi-dimensional control method of the beauty device as described in any one of the above embodiments.

[0264] Figure 18 is a schematic structural diagram of a beauty device provided by an embodiment of the present invention. The beauty device 1800 includes: a processor 180, a memory 181, a bus 182 and a communication interface 183. The processor 180, the communication interface 183 and the memory 181 are connected via the bus 182; the processor 180 is configured to execute an executable module stored in the memory 181, such as a computer program.

[0265] The memory 181 may include high-speed random access memory (RAM) and may also include non-volatile memory, such as at least one disk storage. The system network element communicates with at least one other network element via at least one communication interface 183 (which may be wired or wireless), and may utilize the Internet, a wide area network, a local area network, a metropolitan area network, or the like.

[0266] The bus 182 may be an ISA bus, a PCI bus, or an EISA bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, FIG18 is shown with only one bidirectional arrow, but this does not mean that there is only one bus or one type of bus.

[0267] Among them, the memory 181 is configured to store a program, and the processor 180 executes the program after receiving an execution instruction. The method executed by the device for flow process definition disclosed in any of the aforementioned embodiments of the present invention can be applied to the processor 180 or implemented by the processor 180.

[0268] The processor 180 may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method may be completed by hardware integrated logic circuits or software instructions in the processor 180. The processor 180 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It may implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of the present invention. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in conjunction with the embodiments of the present invention may be directly implemented and executed by a hardware decoding processor, or by a combination of hardware and software modules in the decoding processor. The software module can be located in a storage medium well-known in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, etc. The storage medium is located in memory 181, and processor 180 reads the information in memory 181 and, in conjunction with its hardware, completes the steps of the above method.

[0269] An embodiment of the present invention provides a beauty system, comprising: the beauty device provided in the aforementioned embodiment, and a server connected to the beauty device.

[0270] The computer program product of the readable storage medium provided in the embodiment of the present invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be configured to execute the method described in the previous method embodiment. The specific implementation can be referred to the previous method embodiment and will not be repeated here.

[0271] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0272] Example 4

[0273] Milk light can be configured to form water in the middle to deep layers of the dermis. When the surface layer of water is irradiated by milk light, the distance between each water molecule increases, making it more fluid. The mitochondria in the cells are powered by an enzyme bound to their cell membranes. This enzyme rotates like a molecular turbine, and being surrounded by more fluid water makes it easier to rotate, thereby producing more ATP (Adenosine triphosphate).

[0274] ATP promotes cell repair and regeneration. ATP, or adenosine triphosphate, is a coenzyme and the direct source of energy for all life activities of tissue cells in the body. It promotes cell repair and regeneration, enhances the biological activity of collagen fibers, redistributes the dermis, and repairs damaged skin. This makes the skin an excellent reflector, indirectly increasing the skin's surface refractive index.

[0275] In addition, the photothermal effect can promote the metabolic decomposition of pigments. The photothermal effect produced by milk light can enhance vascular function, dilate microvessels, accelerate blood circulation, increase blood oxygen content, and accelerate the elimination of metabolic products.

[0276] However, milk light beauty devices generally emit light through tungsten filaments, but tungsten filament light has a geometric light divergence mode, and each point is divergent in spherical space. For the optical system control technology, there is a large loss of optical energy and it is converted into heat, which increases the heat dissipation burden of the system. The existing conventional heat dissipation technology mostly adopts single or comprehensive fusion technologies such as heat pipes, TEC and air cooling in parallel, which increases the size of the equipment. Due to the particularity of the application scenario, heat pipe heat dissipation technology cannot achieve high heat dissipation efficiency. Ordinary TEC heat dissipation has the problem that the temperature of the heat dissipation module close to it increases due to the high temperature of tungsten filament light emission and thermal radiation itself. At this time, the temperature difference between the preset management temperature and the actual heat dissipation module temperature is too large. In the TEC effect, the greater the temperature difference, the worse the cooling effect, and the working efficiency of the load 400 is reduced accordingly.

[0277] Based on this, referring to FIG. 19 , the present application provides a photonic beauty device control system, which is applied to a photonic beauty device. The system includes a first processor 100 , a temperature adjustment module 200 , and an alarm module 300 . The temperature adjustment module 200 includes a semiconductor refrigeration unit 210 , a water cooling unit 220 , a radiator 230 , and a fan 240 .

[0278] The first processor 100 is in communication with the temperature adjustment module 200 and is electrically connected to the alarm module 300. The first processor 100 is also electrically connected to the load 400. When the load 400 is in operation, the first processor 100 controls the operation of the temperature adjustment module 200. The first processor 100 is configured to increase the power of the semiconductor refrigeration unit 210, the water cooling unit 220, and the fan 240 when the temperature adjustment module 200 detects that the instrument temperature of the photonic cosmetic device exceeds a temperature threshold. The first processor 100 is further configured to reduce the power of the load 400 until the instrument temperature returns to the temperature threshold if the instrument temperature does not return to the temperature threshold within a first preset time period.

[0279] If the temperature of the instrument does not return to the temperature threshold within the second preset time period after the power of the load 400 is reduced, the load 400 is controlled to be shut down, and the alarm module 300 issues an alarm.

[0280] Specifically, the control process of the photon beauty device control system is described with reference to FIG20, FIG23 and FIG24:

[0281] First, referring to Figure 20, Figure 20 shows a schematic diagram of the pin definition of the first processor 100 provided in this application. Taking the first processor 100 as an STM32 model MCU and the second processor 600 as a CPU or GPU or a combination of the two as an example, Figure 20 shows the pin definition of MCU (Microcontroller Unit, micro control unit) U6, whose specific model can be STM32F103RCT6. The use of dual processors can give full play to their respective advantages, realize parallel work, and reduce the possibility of efficiency brought by serial work.

[0282] The MCU is configured to transmit and receive signals between modules, as well as process them. The CPU / GPU enables efficient local image processing and analysis, even when disconnected from the network. This leverages the real-time response capabilities of image processing, enabling millisecond-level sampling, analysis, comparison, and intervention at the device's output terminals. A communication protocol is established via the serial port between the CPU and MCU (STM32), enabling adjustments to the STM32 using coded instructions.

[0283] The MCU (STM32) core itself has running program code. When receiving instructions from the CPU, it calls the coded instructions to complete the work assigned to the main control. The running program of the MCU (STM32) itself includes the control of devices such as the TEC, motor and load 400. The real-time adjustment of the load 400 is also completed by receiving instructions from the CPU at the same time. The working instructions of the CPU when the user operates the touch screen 700 include the modulation of mode response, power intensity and working status.

[0284] It should be noted that Figure 20 shows the complete pin definition, but in actual applications, not all pins are necessarily used. Figures 21-24 will be discussed later. Signals identical to those in Figure 20 appear in Figures 21-24, indicating that the end containing the signal is electrically connected to the corresponding pin of U6. Furthermore, Figure 20 also illustrates a set of LC filter units, including C20, C21, and L4, which are electrically connected to the MCU's power supply VCC_3V3, forming a filter between the chip and the power supply. Furthermore, ADC_ADN and ADC_TEMP represent analog-to-digital conversion of other types of signals in the circuit, converting other types of signals that were originally analog signals into digital signals.

[0285] For example, the present application can use two semiconductor refrigeration chips, designated TEC-1 and TEC-2, and two NTC thermistors, designated NTC-1 and NTC-2, for collaborative cooling. When load 400 is in standby mode, the semiconductor refrigeration chip or one of the semiconductor refrigeration chips, the motor, and fan 240 continue to operate. After one minute of normal operation, the voltage of the semiconductor refrigeration chip or one of the semiconductor refrigeration chips drops to 5V. If load 400 remains in standby mode for more than two minutes, load 400 is automatically disconnected, and other components (including the semiconductor refrigeration chip, thermistor chip, motor, fan, etc.) are also turned off.

[0286] In an optional embodiment, the system further includes a vibration motor 500, which is electrically connected to the first processor 100. The vibration motor 500 is configured to provide vibration feedback when an image instruction on the touch screen is operated; secondly, when the device is operating, vibration is required to increase the absorption of skin care products by the skin.

[0287] In an optional embodiment, the system further includes a second processor 600, which is communicatively connected to the first processor 100;

[0288] The second processor 600 is also connected to the display screen 700 and the camera 800 for communication.

[0289] In an optional embodiment, the semiconductor refrigeration unit 210 includes a semiconductor refrigeration sheet and an NTC thermistor chip, and the NTC thermistor chip is configured to monitor the temperature of the instrument.

[0290] In an optional embodiment, the temperature adjustment module 200 further includes a TEC control module, the TEC control module is communicatively connected to the first processor 100, and the TEC control module is also communicatively connected to the semiconductor refrigeration plate;

[0291] The TEC control module is configured to send a first cooling signal to the semiconductor refrigeration chip, and the first cooling signal is configured to enable the semiconductor refrigeration chip to adjust the temperature of the front end of the photonic beauty device.

[0292] In an optional embodiment, the TEC control module is further in communication with the water cooling unit 220;

[0293] The TEC control module is further configured to send a second cooling signal to the water cooling unit 220 , wherein the second cooling signal is configured to enable the operation of the water cooling unit 220 to adjust the temperature of the cavity of the photonic beauty device.

[0294] Please refer to Figure 21, which shows the control circuit of TEC, wherein TEC_EN2 is mainly configured as a refrigeration plate to enable the cooling of the water cooling system; TEC_EN3 is mainly configured as a refrigeration plate to enable the cooling part of the front end of the device that contacts the skin, so that the part of the device that contacts the skin is maintained at the set temperature value. TEC_EN2 and TEC_EN3 work together to enable each cooling part to achieve the function of adjusting the temperature.

[0295] In an optional embodiment, the TEC control module includes a first switch tube Q1, a first voltage regulator U1, a second switch tube Q2 and a second voltage regulator U2.

[0296] The control end of the first switching tube is electrically connected to the first processor 100, the output end of the first switching tube is grounded, and the input end of the first switching tube is electrically connected to the output end of the first voltage regulator; the input end of the first voltage regulator is electrically connected to the first processor 100; the control end of the second switching tube is electrically connected to the first processor 100, the output end of the second switching tube is grounded, and the input end of the second switching tube is electrically connected to the output end of the second voltage regulator; the input end of the second voltage regulator is electrically connected to the first processor 100.

[0297] The TEC control circuit also includes terminal H1, configured to connect two circuits; protection resistors R1 and R2 for the first switch Q1, with one end of R2 electrically connected to pin 56 of U6 and configured to receive the MCU's control signal EN_Q2; and protection resistors R8 and R9 for the second switch Q2, with one end of R9 electrically connected to pin 62 of U6 and configured to receive the MCU's control signal EN_Q3. The TEC control circuit also includes a filtering network consisting of L1, C1, C2, C3, C4, R3, R4, and C5. This filtering, through the first voltage regulator U1 and the filtering network, significantly reduces circuit noise. R5 and R6 are output sampling resistors that divide the output signal, which is then current-limited by protection resistor R7 to produce the output signal DAC_VOUTA1. The term "DAC" refers to the subsequent analog-to-digital conversion required. The TEC control circuit also includes a filter network composed of L2, C6, C7, C8, C9, R10, R11, and C10, as well as output sampling resistors R12 and R13 and a current limiting resistor R14. The principles are the same as above and will not be repeated here.

[0298] Referring to FIG. 23 , in one embodiment, the TEC control circuit further includes a third voltage regulator U3. The output of the third voltage regulator U3 is electrically connected to the MCU and configured to output a TEC_EN1 signal. The function of the TEC_EN1 signal is similar to that of TEC_EN2 and TEC_EN3, enabling the cooling system. The enabled device here can be a fan or other cooling device. The branch containing the third voltage regulator U3 also includes a filter network consisting of L3, C12, C13, C14, C15, R16, R17, and C16, as well as output sampling resistors R18 and R19 and a current-limiting resistor R20. The principles are the same as above and will not be further described here. The first and second switching transistors Q1 and Q2 can be NCEP3040Q N-channel superjunction MOS transistors. The first, second, and third voltage regulators U1, U2, and U3 can be SY8368AQQC voltage regulators.

[0299] In an optional embodiment, referring to FIG. 22 , a first pin of the NTC thermistor chip H2 is electrically connected to a low-voltage power supply VCC_3V3, and a second pin of the NTC thermistor chip is communicatively connected to the first processor 100;

[0300] The first processor 100 is configured to obtain the instrument temperature sampled by the NTC thermistor chip, modulate the instrument temperature, and obtain a cooling adjustment instruction;

[0301] The first processor 100 is further configured to adjust the power of the semiconductor refrigeration chip based on the refrigeration adjustment instruction.

[0302] The ADC-ADN is connected to pin 23 of the STM32. After sampling and acquiring temperature data, it modulates the data value on the TEC pin, using voltage regulation to modulate the TEC power. Specifically, the NTC thermistor chip samples the temperature data, converts it through the ADC, and sends the value to the STM32. The chip can read and recognize the value and adjusts the TEC voltage if the temperature is too high or too low. Specifically, if the temperature is too high, the TEC voltage is increased, increasing the cooling capacity and causing the temperature to drop. If the temperature is too low, the TEC voltage is decreased, reducing the cooling capacity and causing the temperature to rise. This calculus-based regulation method completes this temperature maintenance process.

[0303] In addition, the control circuit of the NTC thermistor chip H2 also includes an RC filtering structure composed of a capacitor C11 and a resistor R15, which realizes filtering between the NTC thermistor chip and the MCU.

[0304] In an optional embodiment, the first processor 100 is further configured to control the temperature adjustment module 200 to turn off if the load 400 is in the standby state for a third preset time period;

[0305] For example, the standby time, i.e., the third preset time period, may be within 2 minutes. When the device is on standby for a period of time without any action by the user, the device will automatically shut down to achieve energy saving.

[0306] If the load 400 is in the working state for a fourth preset time period, the first processor 100 is further configured to turn off the load 400 and turn off the temperature adjustment module 200 after a fifth preset time period.

[0307] For example, if the fourth preset time is 10 minutes and the fifth preset time is 30 seconds, when the load 400 has been operating for 10 minutes (excluding standby time) during one operation, the load 400 is turned off, while the semiconductor cooling plate, the fan 240, and the beauty instrument motor continue to operate for 30 seconds before automatically turning off.

[0308] Furthermore, once the device is started and has been running for more than 10 seconds, the semiconductor cooling plate, fan 240, and beauty instrument motor will automatically stop after running for 30 seconds after the power button is pressed. This is to prevent the device from overheating due to excessive operation and to avoid burns to the user.

[0309] Please refer to Figure 24. When the device is started, the beauty instrument motor starts working at the same time. When the load 400 stops working, the beauty instrument motor continues to work for 3-5 seconds. At the same time, the semiconductor refrigeration plate and the fan 240 also work synchronously with the motor. When the load 400 is in standby mode, the semiconductor refrigeration plate, the beauty instrument motor and the fan 240 are still working. Among them, the beauty instrument motor is a motor that supports the functions of the beauty instrument itself. U4 in Figure 24 is a voltage stabilizer or a constant voltage transformer, and U5 is an electronic potentiometer. The ADJ_EN signal sent by the third pin of U4 is a motor enable signal, which is configured to enable the beauty instrument motor. When the ADJ_EN signal enables the beauty instrument motor, the beauty instrument motor operates normally. When the enable signal is stopped, the beauty instrument motor stops running.

[0310] [Corrected 29.07.2024 according to Rule 91] The ADJ_UD signal output from pin 3 of U5 is the motor direction signal, and the ADJ_CS signal output from pin 4 is the motor high-speed start signal. The power supply voltage of U4 is +5V, and the power supply voltage of U5 is VCC_3V3. In addition, the 3V3_ADJ output from the fifth pin of U4 is the operating voltage of the beauty instrument motor. 3V3_ADJ is also electrically connected to terminal CN1. The circuit shown in Figure 24 also includes isolation capacitors C17, C18 and C19 between the isolation power supply and ground, and a protection resistor R21 between U4 and U5.

[0311] It should be noted that the operating voltage of the motor is adjustable. The currently defined operating voltage is 1.2V. As the gear changes, the power of the load 400 increases, and the supply voltage of the motor changes accordingly. The higher the power of the load 400, the greater the operating voltage of the motor. The range of change is 1.2-3.3V. For example, U4 can be a voltage-adjustable low-dropout regulator TLV74312PDBVR.

[0312] In a specific embodiment, when the NTC thermistor chip detects that the temperature is too high, the power of the semiconductor refrigeration chip is increased, and the operating voltage 3V3_ADJ of the beauty instrument motor and the fan 240 is also increased. If the temperature still does not drop or the drop is not obvious after the power is increased, the power of the load 400 is reduced until the temperature returns to the range of the initial setting value.

[0313] When the temperature is too high and the load is reduced by 400 watts for a period of time but still cannot be restored to a safe value, an alarm will be issued and the machine will enter automatic shutdown state.

[0314] The photonic beauty device control system provided in this embodiment has at least the following advantages:

[0315] This embodiment utilizes a combination of water cooling, semiconductor cooling chips, and fans to cool the photonic beauty device, effectively preventing the temperature-dependent and inefficient cooling of a single cooling device, particularly semiconductor cooling chips. Furthermore, if the device overheats, it promptly shuts off power and issues an early warning, ensuring safety during use.

[0316] The present invention provides a photon cosmetic device, comprising the photon cosmetic device control system according to any one of the aforementioned embodiments.

[0317] It should be noted that the beauty instrument, skin beauty instrument, beauty equipment and photon beauty equipment described in the above embodiments 1 to 4 are all different descriptions of the same device and can be understood as the same device.

[0318] In the embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, 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 through some communication interface, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0319] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0320] In addition, each functional unit in the embodiments provided in 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.

[0321] If the functions are implemented in the form of software functional units and sold or used as independent products, they 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 the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0322] It should be noted that similar numbers and letters represent similar items in the following figures. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. In addition, the terms "first", "second", "third", etc. are only configured to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0323] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present application, which are used to illustrate the technical solutions of the present application, rather than to limit them. The scope of protection of the present application is not limited thereto. Although the present application has been described in detail with reference to the above-described embodiments, those skilled in the art should understand that any person skilled in the art can modify or easily conceive of changes to the technical solutions described in the above-described embodiments within the technical scope disclosed in the present application, or perform equivalent replacements for some of the technical features thereof. However, these modifications, changes, or replacements do not deviate from the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application. They should all be included in the scope of protection of the present application. Therefore, the scope of protection of the present application shall be subject to the scope of protection of the claims.

Claims

1. A beauty instrument control method, characterized in that: The method comprises: The touch chip of the beauty instrument is used to identify whether the end surface of the beauty instrument fits the skin; If the end face of the beauty instrument is in contact with the skin, the beauty instrument load is started, and the camera of the beauty instrument is used to detect whether the skin and the end face of the beauty instrument are in relative motion based on motion estimation technology; If the skin and the end face of the beauty instrument do not move relative to each other, the power of the beauty instrument load is reduced until the beauty instrument is in a standby state; and the step of identifying whether the end face of the beauty instrument is in contact with the skin through the touch chip of the beauty instrument is performed again; If the skin and the end surface of the beauty instrument move relative to each other, the beauty instrument is controlled to work normally.

2. The beauty instrument control method according to claim 1, characterized in that: The detecting whether the skin and the end surface of the beauty instrument have relative motion by using the camera of the beauty instrument based on motion estimation technology includes: Acquiring a state image of the fitting surface between the skin and the end surface of the beauty instrument through the camera; Determining whether there is pixel offset in the state image based on the motion estimation technology; If there is a pixel offset in the state image, then the skin and the end surface of the beauty instrument undergo relative motion; If there is no pixel offset in the state image, then there is no relative movement between the skin and the end surface of the beauty instrument.

3. The beauty instrument control method according to claim 2, characterized in that: The determining whether there is pixel offset in the state image based on the motion estimation technology includes: Calculate the displacement vectors corresponding to each pixel point in the state image of the current frame and each pixel point in the state image of the previous frame by using the motion estimation technology; Determine whether the displacement vector corresponding to each of the pixel points is greater than a displacement threshold; If there is a displacement vector greater than the displacement threshold, then there is a pixel offset in the state image; If there is no displacement vector greater than the displacement threshold, then there is no pixel offset in the status image.

4. The beauty instrument control method according to claim 1, characterized in that: The controlling the beauty instrument to operate normally includes: Detecting the relative motion speed of the end face of the beauty instrument based on the motion estimation technology; Outputting a target level signal according to the relative motion speed; adjusting the power of the beauty instrument load according to the target level signal; The beauty instrument is controlled to work according to the power of the beauty instrument load.

5. The beauty instrument control method according to claim 1, characterized in that: After the beauty instrument is controlled to work normally, the following steps are included: Detecting the end surface temperature of the end surface of the beauty instrument, and determining the temperature change of the end surface temperature; If the temperature change is an increase, the power of the cooling plate of the beauty instrument is increased through a control algorithm; If the temperature change is unchanged, the power of the refrigeration plate is controlled to remain unchanged; If the temperature change is a decrease, the power of the cooling plate of the beauty instrument is reduced through a control algorithm.

6. A beauty instrument control device, characterized in that: The device comprises: An identification module configured to identify whether the end face of the beauty instrument is in contact with the skin through a touch chip of the beauty instrument; a detection module configured to start a beauty instrument load if the end face of the beauty instrument is in contact with the skin, and detect whether the skin and the end face of the beauty instrument are in relative motion through a camera based on motion estimation technology; The first control module is configured to reduce the power of the load until the beauty instrument is in a standby state if the skin and the end face of the beauty instrument do not move relative to each other; and re-execute the step of identifying whether the end face of the beauty instrument is in contact with the skin through the touch chip of the beauty instrument; The second control module is configured to control the beauty instrument to work normally if the skin and the end surface of the beauty instrument move relative to each other.

7. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the steps of the beauty instrument control method according to any one of claims 1 to 5 are implemented.

8. A computer-readable storage medium, characterized in that: It stores a computer program, which, when executed by a processor, implements the steps of the beauty instrument control method according to any one of claims 1 to 5.

9. A skin data processing method, characterized in that: Applicable to a skin beauty instrument with a high-speed multi-spectral imaging module at the front end; the skin data processing method includes: Acquire the current detailed skin image collected by the high-speed multispectral imaging module; Determining the current skin position of the skin beauty instrument by comparing the current detailed skin image with a pre-collected full-face skin image; According to the current detailed skin image and the current skin position, current treatment parameters of the skin beauty instrument are determined.

10. The skin data processing method according to claim 9, characterized in that: The step of comparing the current detailed skin image with a pre-collected full-face skin image to determine the current skin position of the skin beauty device comprises: Obtaining a current facial pattern code corresponding to the current detailed skin image; Matching the current facial pattern code with a plurality of original facial pattern codes corresponding to the full-face skin image; The skin position corresponding to the successfully matched original facial pattern code is determined as the current skin position of the skin beauty device.

11. The skin data processing method according to claim 10, characterized in that: The current facial pattern code includes a feature vector; the step of obtaining the current facial pattern code corresponding to the current detailed skin image includes: Acquire current skin feature data in the current detailed skin image; wherein the current skin feature data includes attribute data under preset parameters, the preset parameters include one or more of pores, spots and textures, and the attributes corresponding to the attribute data include one or more of color, size, position, distance and area; A feature vector corresponding to the current detailed skin image is generated according to the current skin feature data.

12. The skin data processing method according to claim 9, characterized in that: The step of determining the current treatment parameters of the skin beauty instrument according to the current detailed skin image and the current skin position includes: Acquire the last detailed skin image and the last treatment parameter corresponding to the current skin position; wherein the last detailed skin image is a detailed skin image acquired by the high-speed multispectral imaging module at the current skin position before the most recent treatment, and the last treatment parameter is a treatment parameter corresponding to the current skin position during the most recent treatment; Determining a previous treatment effect according to the current detailed skin image and the previous detailed skin image; According to the last treatment effect and the last treatment parameter, the current treatment parameter of the skin beauty instrument is determined.

13. The skin data processing method according to claim 12, characterized in that: The determining the last treatment effect according to the current detailed skin image and the last detailed skin image comprises: Acquire current skin feature data in the current detailed skin image and previous skin feature data in the previous detailed skin image; Comparing the current skin feature data with the previous skin feature data to obtain a comparison result; The last treatment effect is determined according to the comparison result.

14. The skin data processing method according to claim 12, characterized in that: The last treatment effect includes a positive change, no change or a negative change; and determining the current treatment parameter of the skin beauty instrument according to the last treatment effect and the last treatment parameter includes: When the last treatment effect is a positive change, determining the last treatment parameter as the current treatment parameter of the skin beauty instrument; When the last treatment effect is unchanged, the treatment parameter obtained by increasing the last treatment parameter by a first preset ratio is determined as the current treatment parameter of the skin beauty instrument; When the last treatment effect is a negative change, the treatment parameter obtained by reducing the last treatment parameter by a second preset ratio is determined as the current treatment parameter of the skin beauty instrument.

15. The skin data processing method according to claim 9, characterized in that: The step of determining the current treatment parameters of the skin beauty instrument according to the current detailed skin image and the current skin position includes: Obtaining preset basic treatment parameters and current skin impedance value corresponding to the current skin position; The preset basic treatment parameters are adjusted according to the current skin impedance value to obtain the current treatment parameters of the skin beauty instrument.

16. The skin data processing method according to claim 9, characterized in that: The step of determining the current treatment parameters of the skin beauty instrument according to the current detailed skin image and the current skin position includes: The current detailed skin image and the current skin position are input into a data model obtained based on deep learning, and the treatment parameters output by the data model are used as the current treatment parameters of the skin beauty device; wherein the data model is trained based on historical detailed skin images and historical treatment parameters corresponding to different skin positions of the current user.

17. A skin data processing device, characterized in that: Applicable to a skin beauty instrument with a high-speed multi-spectral imaging module at the front end; the skin data processing device comprises: An image acquisition module, configured to acquire a current detailed skin image acquired by the high-speed multispectral imaging module; a position determination module, configured to determine the current skin position of the skin beauty instrument by comparing the current detailed skin image with a pre-collected full-face skin image; The parameter determination module is configured to determine the current treatment parameters of the skin beauty instrument according to the current detailed skin image and the current skin position.

18. A skin beauty instrument, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and when the processor executes the computer program, the skin data processing method according to any one of claims 9 to 16 is implemented.

19. A skin beauty system, characterized in that: The skin beauty instrument comprises the skin beauty instrument as claimed in claim 18, and further comprises a user terminal and a cloud server respectively connected to the skin beauty instrument.

20. A multi-dimensional control method for beauty equipment, characterized in that: The method is applied to a server and comprises: Obtain electrode signals, skin sampling data and image data collected by beauty equipment; determining whether the beauty device is in contact with the skin of the current user based on the electrode signal; If the skin of the current user is in contact, determining the working state of the beauty device based on the skin sampling data; The skin condition of the current user is determined based on a pre-trained skin detection model and the image data.

21. The multi-dimensional control method of the beauty device according to claim 20, characterized in that: Determining the skin condition of the current user based on a pre-trained skin detection model and the image data includes: The image data is input into a pre-trained skin detection model, and the skin type and the points of the skin type in the image data are labeled to obtain the skin condition of the current user; wherein the skin condition includes: the skin type and the points corresponding to the skin type, and the skin type includes one or more of the following: normal skin, acne, wrinkles and spots.

22. The multi-dimensional control method of the beauty device according to claim 20, characterized in that: The training of the skin detection model includes: Acquire training images; wherein the training images include: marked skin images and unmarked skin images; Performing cross-domain image style transfer on the training image; Prototype network training is performed based on labeled skin images after cross-domain image style transfer to obtain category prototypes; Using a cluster center nearest neighbor rule, the unlabeled skin image is matched with the category prototype to generate a pseudo label for the unlabeled skin image; Based on the pseudo-label of the unlabeled skin image, the skin detection model is trained to obtain a trained skin detection model.

23. The multi-dimensional control method of the beauty device according to claim 20, characterized in that: Get training images, including: Get initial training images; Inputting the initial training image into a feature extractor to extract features and obtain a feature image; The feature image is input into an image generator for image reconstruction to obtain a training image.

24. A multi-dimensional control device for beauty equipment, characterized in that: The device is applied to a server and includes: A data acquisition module configured to acquire electrode signals, skin sampling data, and image data collected by the beauty device; an electrode signal determination module, configured to determine whether the beauty device is in contact with the skin of a current user based on the electrode signal; a working state determination module configured to determine the working state of the beauty device based on the skin sampling data if the skin of the current user is in contact; The skin condition determination module determines the skin condition of the current user based on a pre-trained skin detection model and the image data.

25. A beauty system, characterized in that: include: The beauty device also includes a server connected to the beauty device.

26. A photon beauty equipment control system, characterized in that: Applied to photon beauty equipment, the system includes a first processor, a temperature adjustment module and an alarm module, and the temperature adjustment module includes a semiconductor refrigeration unit, a water cooling unit, a radiator and a fan; The first processor is communicatively connected to the temperature adjustment module, and the first processor is electrically connected to the alarm module; The first processor is also electrically connected to the load, and when the load is in a working state, the first processor controls the temperature adjustment module to operate; The first processor is configured to increase the power of the semiconductor refrigeration unit, the water cooling unit and the fan when the temperature adjustment module detects that the instrument temperature of the photon beauty device exceeds a temperature threshold; The first processor is further configured to reduce the power of the load until the temperature of the instrument drops back to the temperature threshold if the temperature of the instrument does not drop back to the temperature threshold within a first preset time period; If the temperature of the instrument has not dropped back to the temperature threshold within a second preset time after the load power is reduced, the load is controlled to shut down and the alarm module sounds an alarm.

27. The photon beauty device control system according to claim 26, characterized in that: The first processor is further configured to control the temperature adjustment module to shut down if the load is in the standby state for a third preset time period; If the load is in the working state for a fourth preset time period, the first processor is further configured to turn off the load, and turn off the temperature adjustment module after a fifth preset time period.

28. The photon beauty device control system according to claim 26, characterized in that: The semiconductor refrigeration unit includes a semiconductor refrigeration sheet and an NTC thermistor chip, and the NTC thermistor chip is configured to monitor the temperature of the instrument.

29. The photon beauty device control system according to claim 28, characterized in that: The temperature adjustment module further includes a TEC control module, the TEC control module is communicatively connected to the first processor, and the TEC control module is also communicatively connected to the semiconductor refrigeration sheet; The TEC control module is configured to send a first cooling signal to the semiconductor cooling chip, and the first cooling signal is configured to enable the semiconductor cooling chip to adjust the temperature of the front end of the photon beauty device.

30. The photon beauty device control system according to claim 29, characterized in that: The TEC control module is also in communication connection with the water cooling unit; The TEC control module is further configured to send a second cooling signal to the water cooling unit, wherein the second cooling signal is configured to enable the operation of the water cooling unit to adjust the temperature of the cavity of the photon beauty device.

31. The photon beauty device control system according to claim 29, characterized in that: The first pin of the NTC thermistor chip is electrically connected to the low-voltage power supply, and the second pin of the NTC thermistor chip is communicatively connected to the first processor; The first processor is configured to, after acquiring the instrument temperature sampled by the NTC thermistor chip, modulate the instrument temperature to obtain a refrigeration adjustment instruction; The first processor is further configured to adjust the power of the semiconductor refrigeration chip based on the refrigeration adjustment instruction.

32. A photon beauty device, characterized in that: The photon beauty device comprises the photon beauty device control system as described in any one of claims 26-31.

Images