Search

CN-121987337-A - Hair follicle positioning dehairing device and method based on multi-frequency skin electrical impedance scanning

CN121987337ACN 121987337 ACN121987337 ACN 121987337ACN-121987337-A

Abstract

The invention discloses a hair follicle positioning dehairing device and method based on multi-frequency skin electrical impedance scanning, which belong to the technical field of medical cosmetology, wherein the device comprises six modules of microelectrode array, impedance measurement, visual detection, control, laser control and display prompt, all the modules are cooperatively matched to realize the integral functions of skin electrical signal transmission, impedance data acquisition and calculation, hair follicle visual detection, bimodal fusion positioning, laser precise irradiation and device working and skin contact state prompt, the method comprises five steps of initializing and calibrating the microelectrode array, acquiring S2 impedance signals, constructing and filtering the S3 impedance spectrum, detecting the candidate position of the hair follicle, fusing visual and impedance detection results to finish the accurate positioning and output of the hair follicle, and the invention adopts impedance and visual bimodal fusion positioning, eliminates contact impedance interference by combining a four-electrode measurement method, improves the hair follicle positioning precision through multi-step data optimization and refinement treatment, and has high practicability.

Inventors

  • CHEN ZIYI

Assignees

  • 壹万万(广州)科技有限公司

Dates

Publication Date
20260508
Application Date
20260330

Claims (10)

  1. 1. The hair follicle positioning dehairing device based on the multi-frequency skin electrical impedance scanning is characterized by comprising a microelectrode array module, an impedance measuring module, a visual detection module, a control module, a laser control module and a display prompting module; the microelectrode array module is electrically connected with the impedance measurement module through the flexible circuit board and is used for contacting skin and realizing electric signal transmission; The impedance measurement module is in communication connection with the control module and is used for generating alternating current excitation signals, performing electrode gating and collecting and calculating local impedance data of the skin; the visual detection module is in communication connection with the control module and is used for collecting skin images, detecting hair follicle positions and outputting a visual hair follicle position set; The control module is in communication connection with each module to realize cooperative control, and is used for receiving impedance data of the impedance measurement module and hair follicle position data of the visual detection module, constructing an impedance map according to the impedance data, determining hair follicle candidate positions, performing space matching and fusion positioning on the hair follicle candidate positions and detection results of the visual detection module, and outputting a fused hair follicle coordinate set; The laser control module is in communication connection with the control module and is used for executing laser irradiation according to the fused hair follicle coordinate set output by the control module; the display prompt module is electrically connected with the control module and is used for sending out skin contact state prompts and device working state prompts.
  2. 2. The hair follicle-localized hair removal device based on multi-frequency electrical skin impedance scanning of claim 1, wherein the microelectrode array module comprises a 4 x 4 = 16 microelectrode array with an electrode spacing of between 1 and 3mm and a diameter of between 0.5 and 0.8mm, and wherein the microelectrode array module has gold-plated copper as the electrode material.
  3. 3. The hair follicle positioning dehairing device based on multi-frequency skin electrical impedance scanning according to claim 1, wherein the impedance measuring module comprises a signal generating unit, a signal collecting unit and an electrode selection communicating unit, wherein the signal generating unit is used for generating a multi-frequency sinusoidal alternating current excitation signal, the signal collecting unit is used for collecting a voltage signal and combining the excitation signal to complete local impedance calculation, the electrode selection communicating unit is used for realizing electrode selection and passage communication of a microelectrode array, the excitation current output by the signal generating unit is less than or equal to 1mA, and the excitation frequency of the multi-frequency sinusoidal alternating current excitation signal generated by the signal generating unit is 10k/100k/1 MHz.
  4. 4. The hair follicle positioning dehairing device based on multi-frequency skin electrical impedance scanning as recited in claim 3, wherein the signal generating unit and the signal acquisition unit are integrated on an AFE chip, the electrode selection communication unit is composed of an analog multiplexer, the control module is an MCU, the MCU is in communication connection with the AFE chip of the impedance measurement module and the analog multiplexer through an SPI bus, the impedance measurement module distributes the paths of the electrodes through the electrode selection communication unit, current injection and voltage acquisition path separation of a four-electrode measurement method are achieved, and interference of contact impedance between the electrodes and skin is eliminated.
  5. 5. A hair follicle positioning method based on multi-frequency skin electrical impedance scanning, which is applied to the dehairing device according to any one of claims 1-4, and is characterized by comprising the following steps: S1, initializing and calibrating a microelectrode array, namely completing power-on self-detection, electrode connectivity detection, air baseline acquisition and skin contact detection of the microelectrode array through a control module, and providing a reference for subsequent impedance measurement; S2, collecting multi-frequency impedance signals, namely, applying multi-frequency alternating current excitation signals with preset parameters to the microelectrode array which is calibrated and effectively contacted with the skin in the step S1 through an impedance measurement module, completing full-array impedance scanning through a four-electrode measurement method, and collecting local impedance data under multiple frequencies; S3, constructing and filtering an impedance spectrum, namely generating an impedance distribution spectrum through interpolation according to the multi-frequency local impedance data acquired in the step S2, and then obtaining low-frequency, medium-frequency and high-frequency impedance characteristic spectrums through Gaussian smoothing denoising treatment; S4, detecting hair follicle candidate positions, namely sequentially completing local minimum value searching, threshold value screening and sub-pixel refinement on the low-frequency impedance characteristic spectrum in the step S3, and performing multi-frequency cross validation to obtain a hair follicle candidate position set with confidence; S5, fusing the visual and impedance detection results to finish the accurate positioning and output of the hair follicle, namely, performing coordinate system alignment, space matching and result fusion on the hair follicle candidate position set obtained in the step S4 and the hair follicle position set output by the visual detection module, and outputting the fused hair follicle coordinate set to the laser control module to execute laser irradiation.
  6. 6. The method for locating hair follicles based on multi-frequency skin electrical impedance scanning as claimed in claim 5, wherein step S1 comprises the steps of: s101, powering on and self-checking, namely after the device is started, a control module sends an initialization instruction to an impedance measurement module, and the impedance measurement module carries out self-calibration on a signal generation unit and a signal acquisition unit in the impedance measurement module by configuring gain, excitation frequency and sampling rate parameters of impedance measurement; S102 electrode connectivity detection, wherein a control module sequentially selects and communicates each electrode of a microelectrode array through an electrode selection communication unit, applies a 100kHz/100 mu A test signal, measures impedance values from each electrode to a common reference electrode, marks as a fault electrode if the electrode open circuit impedance is more than 10MΩ, and skips in subsequent calculation; s103, acquiring an air baseline, namely finishing one-time microelectrode array full-array scanning through an impedance measurement module when the device is not contacted with skin, and recording a baseline impedance value Z_air of each electrode pair in the air; s104, skin contact detection, namely after the device is attached to the skin, the impedance measuring module detects the electrode impedance value in real time, when the electrode pair impedance value exceeding 75% is reduced to less than 1/10 of Z_air, the skin is judged to be in effective contact, a measuring mode is entered, and otherwise, an attaching adjustment prompt is sent out through the display prompt module.
  7. 7. The method of hair follicle localization based on multi-frequency skin impedance scanning of claim 6, wherein step S2 comprises the steps of: S201, configuring excitation signals, namely controlling a signal generating unit to generate sinusoidal excitation signals of 10kHz, 100kHz and 1MHz by a control module, wherein excitation current is in a constant current mode, the amplitude is less than or equal to 1mA, and the peak value Vpp of excitation voltage is less than 1V; S202 four-electrode measuring method is implemented by selecting four adjacent electrodes from a microelectrode array to form a measuring unit, wherein two electrodes at two sides are current injection electrodes, exciting current Iexc is applied, two electrodes in the middle are voltage acquisition electrodes, measuring potential difference Vbc, and finally, the method is carried out according to the following steps of Calculating local impedance; S203, a full array scanning sequence, namely sequentially selecting and communicating four electrode groups in horizontal, vertical and diagonal directions on the microelectrode array through an electrode selection communicating unit, wherein each group of four electrodes respectively completes impedance measurement at three frequencies, and the single measurement time is 0.5-1ms; And S204, recording data of a data set { Z (i, j, f_k) }, i, j of the full array scanning through the control module, wherein k=1, 2,3 is a frequency index, f_k is a frequency value, and the data set is temporarily stored in a storage unit of the control module.
  8. 8. The method for locating hair follicle based on multi-frequency electrical impedance scanning of skin of claim 7, wherein step S3 comprises the steps of, S301, interpolation of original data, namely mapping discrete local impedance measured values to a regular grid by adopting a bilinear interpolation method, and respectively generating impedance distribution maps Z10k (x, y), Z100k (x, y) and Z1M (x, y) for three frequencies of 10kHz, 100kHz and 1 MHz; s302 Gaussian smoothing denoising, namely applying a two-dimensional Gaussian filter to each impedance distribution map, enabling sigma of a Gaussian kernel function to obtain electrode spacing which is 0.5-1 times, eliminating random noise, and retaining effective signals of hair follicle scales to obtain a low-frequency impedance characteristic map Z10k_smooth (x, y), a medium-frequency impedance characteristic map Z100deg.K_smooth (x, y) and a high-frequency impedance characteristic map Z1Mk_smooth (x, y); S303, cole-Cole parameter extraction, namely, fitting a Cole-Cole model to each spatial position (x, y) by utilizing impedance characteristic maps of three frequencies, extracting low-frequency impedance R_0, high-frequency impedance R_inf and characteristic time constant tau, wherein a dispersion coefficient alpha takes a fixed empirical value of 0.8 to distinguish a hair follicle region from a non-hair follicle region, the tau value of the hair follicle region is smaller than that of the non-hair follicle region, and the alpha value is closer to 1.
  9. 9. The method for locating hair follicles based on multi-frequency skin electrical impedance scanning as claimed in claim 8, wherein step S4 includes the steps of: S401, searching local minimum values, namely traversing each pixel point on the smoothed low-frequency impedance characteristic spectrum Z10k_smooth (x, y), comparing the impedance value of each pixel point with the impedance values of 8 surrounding adjacent pixel points, and marking the pixel points with the impedance values smaller than 8 adjacent pixel points as local minimum value candidate points; S402, threshold value screening, namely calculating the Ratio of the impedance value of each local minimum candidate point to the average value of the impedance values of 8 adjacent pixel points, and confirming the Ratio as the hair follicle candidate position when the Ratio is less than 0.85; S403, sub-pixel positioning refinement, namely, performing sub-pixel positioning on the candidate positions of the hair follicle screened by the threshold value by adopting two-dimensional parabolic fit in a 3X 3 window which takes the candidate positions as the center and comprises 8 adjacent pixel points of the candidate positions of the hair follicle, so as to improve the spatial resolution of the hair follicle positioning; S404 multi-frequency cross verification, namely repeatedly executing S401-S403 to obtain the positions of hair follicle candidate points in each frequency band for impedance feature maps of three frequency bands, respectively, cross-comparing the candidate points detected by the three frequency impedance feature maps, marking the candidate points which are detected in at least two frequency impedance feature maps and have the space distance between the two candidate points being less than 1 electrode distance as high confidence, marking the detected candidate points in only one frequency map as middle confidence, and outputting a hair follicle candidate position set I_set= { (x_i, y_i, confidence_i) }, wherein the confidence value is high or middle.
  10. 10. The method of locating hair follicles based on multi-frequency skin electrical impedance scanning as claimed in claim 9, wherein step S5 includes: And S501, aligning a coordinate system, namely calibrating the relative position relation between the view field of the camera and the microelectrode array in the device manufacturing stage to obtain an affine transformation matrix M_ calib, and converting the corresponding coordinate of the hair follicle candidate position set obtained by impedance detection in the step S4 into an image coordinate during operation: imp is impedance detection, img is image detection; S502, performing space matching, namely performing space adjacent matching on an impedance hair follicle candidate position set I_set converted into image coordinates and a visual hair follicle candidate position set V_set, wherein the matching threshold is 1mm, searching a visual hair follicle point closest to each candidate point in the I_set in the V_set, and judging that the matching pair is the bimodal detection of the same hair follicle when the distance between the two points is smaller than the matching threshold, and obtaining three types of results of bimodal matching, visual detection only and impedance detection only after matching; S503, fusing results, namely firstly mapping qualitative confidence degrees of high and medium marked in the S4 into quantitative values, wherein the high confidence degree=0.7, the middle confidence degree=0.5, the visual detection module outputs 0-1 quantitative confidence degrees, and then fusing coordinates and the confidence degrees, wherein a calculation formula about fused coordinates and the fused confidence degrees is as follows: , , the fusion confidence bimodal rewards 0.1, the upper limit is 1.0, and the weight calculation in the fusion coordinate formula: , The laser irradiation energy is standard energy, V_j and I_i are hair follicle candidate point coordinates output by the vision and impedance module, and w_v/w_i is fusion weight of corresponding coordinates; For the detection result of only the impedance, the confidence level is set to be 0.5-0.7 according to the confidence level of the detection result of only the impedance, 0.7 is taken according to the confidence level of the impedance candidate point, 0.5 is taken according to the high level, and the laser irradiation energy is 70% of the standard energy; S504 is output to a laser control module, wherein a fusion hair follicle set is generated through the control module, wherein F_set { (x_k, y_k, conf_k, energy_k) }, x_k and y_k are coordinates of a kth hair follicle, conf_k is confidence of the kth hair follicle, energy_k is laser irradiation energy of the kth hair follicle, and the fusion hair follicle set is sequenced from high to low according to the confidence and then transmitted to the laser control module.

Description

Hair follicle positioning dehairing device and method based on multi-frequency skin electrical impedance scanning Technical Field The invention relates to the technical field of medical cosmetology, in particular to a hair follicle positioning dehairing device and method based on multi-frequency skin electrical impedance scanning. Background The core of the laser dehairing technology is to precisely position the hair follicle so as to ensure that the laser energy effectively acts on the root of the hair follicle to realize the dehairing effect, and simultaneously, the damage to surrounding normal skin tissues is reduced to the greatest extent. In the prior art, the hair follicle positioning is mainly based on a visual detection method, skin images are acquired through a camera, and then the hair follicle is detected by means of an image recognition algorithm, but the hair follicle is greatly influenced by the surface state of the skin, the positioning accuracy is limited, and the problems of missed detection and false detection are easy to occur. On one hand, the light-colored hair and skin have extremely low optical contrast, the recognition accuracy of a visual model on dark-colored hair can reach more than 95%, the recognition accuracy on light-colored hair such as golden hair, white hair and gray hair is only 60-70%, the missing detection condition is prominent, the problem that light-colored body hair exists in about 30% of population worldwide becomes obvious market dead zone, on the other hand, the phenomenon that black hair root residues exist on the skin surface after shaving and hair follicles exist but no visible hair is exposed exists on the skin surface exists, the pure visual scheme can not detect the hair growing under the skin, and in addition, the visual recognition can be misjudged due to the influence of light conditions such as skin sweat reflection, skin line interference and the like. Hair follicles act as skin appendages, are internally rich in cells, blood vessels and are sufficiently hydrated, and this tissue structure makes a significant difference in electrical properties from the surrounding stratum corneum, dermal connective tissue, particularly in areas of hair follicles where electrical impedance values are typically lower than in surrounding skin areas without hair follicles, which difference can be detected and captured by high sensitivity microelectrode arrays. Based on the above, the existing partial proposal adopts skin impedance measurement as an auxiliary means for hair follicle positioning, but most of the proposal adopts single frequency impedance measurement, and cannot fully reflect the electrical characteristic difference of different tissues of skin, so that the impedance identification degree of hair follicle and surrounding skin is lower, and meanwhile, the conventional two-electrode measurement method is easily interfered by the contact impedance of electrode skin, further increases positioning error and finally influences the effect and the use safety of laser dehairing. Disclosure of Invention Aiming at the technical problems, the invention provides a hair follicle positioning dehairing device and method based on multi-frequency skin electrical impedance scanning, which improve the electrical identification of hair follicles through multi-frequency impedance feature analysis, eliminate contact impedance interference by combining a special measuring method, improve the overall positioning precision through vision matching impedance bimodal fusion positioning, adapt to the requirements of different use scenes and give consideration to the positioning precision and use instantaneity. The invention adopts the following technical scheme for realizing the technical purpose: a hair follicle positioning dehairing device based on multi-frequency skin electrical impedance scanning comprises a microelectrode array module, an impedance measuring module, a visual detection module, a control module, a laser control module and a display prompting module; the microelectrode array module is electrically connected with the impedance measurement module through the flexible circuit board and is used for contacting skin and realizing electric signal transmission; The impedance measurement module is in communication connection with the control module and is used for generating alternating current excitation signals, performing electrode gating and collecting and calculating local impedance data of the skin; the visual detection module is in communication connection with the control module and is used for collecting skin images, detecting hair follicle positions and outputting a visual hair follicle position set; The control module is in communication connection with each module to realize cooperative control, and is used for receiving impedance data of the impedance measurement module and hair follicle position data of the visual detection module, constructing an impedance map acc