Intelligent Scalp Detection System and Its Detection Method

Abstract

The present invention discloses an intelligent scalp detection system and its detection method, which relates to the technical field of intelligent scalp detection. By performing real-time environmental perception and quality detection to obtain shooting conditions and clarity evaluation, using deep learning segmentation and directional enhancement to solve hair occlusion and image noise, leveraging multi-modal fusion and stitching to display the deep structure of the scalp, and outputting diagnosis and optimization strategies in comprehensive analysis and adaptive closed-loop feedback; through software algorithm-level expansion, online training of multi-user long-term data and multi-label pathology expansion can be carried out to achieve adaptive evolution and population statistical analysis; it can effectively improve the stability of image acquisition and the accuracy of segmentation, enhance the adaptability of detection to complex lighting and multi-band information, reduce the risk of misjudgment of pathologies such as grease, inflammation, and fungi, and support personalized care and a wide range of health management applications.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent scalp detection, and specifically to an intelligent scalp detection system and its detection method. Background Art

[0002] In daily life, many users hope to know their scalp health status at any time and place, so as to timely intervene in potential problems such as dandruff, sebum secretion imbalance, hair follicle blockage, and even fungal infections. In this process, users may use a handheld or wearable intelligent detection device and cooperate with a macro camera to take a close-up photo of the scalp, so as to obtain relatively detailed image data. The actual use scenarios include various indoor and outdoor environments such as homes, barbershops, and clinics, and the lighting conditions range from dim to strong direct sunlight; in addition, there are also significant differences in hair density, color, and curl among some people, which makes the scalp area more likely to be blocked by hair, and users may move quickly or deviate in angle during combing or shooting, further increasing the difficulty of image acquisition. Facing these complex and variable environmental factors, how to ensure both the clarity and integrity of scalp images and also take into account usability has become an important challenge for intelligent scalp detection systems.

[0003] After retrieval, in a Chinese invention patent with the authorized announcement number CN113761974B, a method for monitoring the scalp is disclosed, which is applied to an intelligent hair dryer including a photographing device, and includes: when the intelligent hair dryer is running, using the photographing device to photograph the user's head to obtain multiple head photos; identifying the head regions corresponding to the multiple head photos respectively; identifying relevant parameters of each head region based on the head photo corresponding to each head region; and analyzing the scalp condition of each head region based on the relevant parameters of each head region and the historical relevant parameters corresponding to each head region.

[0004] Combined with the above application scenarios and existing technologies:

[0005] Since the reliability of subsequent algorithm analysis needs to be ensured during the scalp acquisition process, when the external light is poor or the focal length cannot be stabilized, the resolution and contrast of scalp images are often difficult to meet the standards, resulting in obvious deviations in recognition or diagnosis results.

[0006] Specifically, if the camera fails to focus sufficiently on the scalp area or is subject to improper user operations (such as rapid movement, obstruction by long curly hair, etc.), problems such as blurred images, distorted edges, or loss of hair follicle details are very likely to occur; this will not only cause a sharp drop in the accuracy of automatic detection of oil concentration, hair follicle density, and minor inflammation, but may also misjudge more serious disease risks such as potential fungal infections and sebaceous gland dysfunction, resulting in users missing the best intervention opportunity or receiving incorrect care advice. From the perspective of industrial application and user experience, insufficient acquisition quality in the above scenarios will directly affect the effectiveness of subsequent multimodal fusion and feature analysis, increase the error rate of algorithm discrimination, and weaken the objective reflection of the actual scalp condition, thus failing to meet the needs of refined detection and personalized health management.

[0007] Therefore, the present invention provides an intelligent scalp detection system and its detection method. Summary of the Invention

[0008] (I) Technical problems to be solved

[0009] In view of the deficiencies of the prior art, the present invention provides an intelligent scalp detection system and its detection method. By real-time environmental perception and quality detection, the shooting conditions and clarity evaluation are obtained. Deep learning segmentation and directional enhancement are used to solve hair occlusion and image noise. Multimodal fusion and stitching are used to display the deep structure of the scalp. And diagnostic and optimization strategies are output in comprehensive analysis and adaptive closed-loop feedback. Finally, through software algorithm level expansion, online training and multi-label pathology expansion can be carried out on long-term data of multiple users, realizing adaptive evolution and population statistical analysis. This solution effectively improves the stability of image acquisition and the accuracy of segmentation, enhances the adaptability of detection to complex lighting and multi-band information, and reduces the risk of misjudgment of pathologies such as oil, inflammation, and fungi, thus solving the technical problems recorded in the background art.

[0010] (II) Technical solutions

[0011] To achieve the above objectives, the present invention is realized through the following technical solutions:

[0012] An intelligent scalp detection method, including that when it is detected that the light intensity and distance meet the available threshold, the environmental perception module combines with the image quality detection unit to calculate the mapped environmental score in real time and output the environmental and preliminary quality data;

[0013] When the environmental and preliminary quality data indicate that imaging is available and the mapped environmental score is higher than the corresponding threshold, the deep learning segmentation model performs hair and scalp separation on the captured image and removes the noise area, calls the PDE-style enhancement component for denoising and super-resolution correction, and outputs the segmented and enhanced image data;

[0014] If the scalp features are stably presented in the segmented and enhanced image data, the multimodal fusion module reads the visible light and near-infrared images, matches the overlapping regions through optimal transformation and performs local compensation to generate multimodal fusion image data;

[0015] Based on the environmental and preliminary quality data and the segmentation results, the intelligent analysis model calls the global health score to identify pathological abnormalities and outputs the analysis result data. At the same time, it transmits back the low-confidence regions for the previous segmentation and environment adaptive optimization iteration;

[0016] When the user generates a large amount of analysis result data during long-term use and expands pathological recognition, multi-label detection is combined to update the model parameters by federated learning, and the global distribution function is used to reveal the coupling between fungal or sebaceous gland abnormalities and environmental factors at the population level, and corresponding treatment strategies are pushed to the user.

[0017] Preferably, the environmental perception module collects the light intensity, the distance value of the camera from the scalp surface, the tilt angle value of the camera relative to the scalp normal direction, and the motion smoothness, and thus defines a mapped environmental score to reflect the usability of the shooting environment.

[0018] Preferably, based on the received mapped environmental score and the initial captured image, the image quality detection unit calculates the image quality score using gradient distribution or wavelet features; the quality score and the mapped environmental score are encapsulated as environmental and preliminary quality data, which also includes environmental illumination, focal length, motion smoothness, and image sharpness; when both the quality score and the mapped environmental score are lower than the corresponding preset thresholds, the user prompt mechanism is triggered.

[0019] Preferably, using the initial captured image as the image input, a pre-trained deep learning segmentation network is used to distinguish the pixels of hair and scalp, combines the pixel-level segmentation result of deep learning with the environmental quality feedback, defines the objective functional and obtains the segmentation mask after solution, identifies the scalp area and non-scalp area according to the segmentation mask, and adaptively smooths or corrects the segmentation edge.

[0020] Preferably, for the scalp area determined by the segmentation mask, a directional enhancement model combining partial differential equations and non-linear response functions is introduced to output the segmented and enhanced image data.

[0021] Preferably, collect multimodal additional image data, including near-infrared bands, multispectral channels, or other-angle shootings;

[0022] Precisely align the segmented and enhanced image data with the multimodal additional image data in the same spatial coordinate system, construct a regional matching energy functional and solve for the optimal transformation:

[0023] After applying the optimal transformation to the multi-modal additional image data, the segmented and enhanced image data and the aligned multi-modal image data in the same coordinate system are obtained.

[0024] Preferably, the segmented and enhanced image data and the multi-modal additional image data are weighted and superimposed with adaptive weights in the same coordinate system to generate multi-modal fusion image data, including the matching results of visible light, near-infrared, and multi-spectral bands.

[0025] Preferably, the intelligent analysis model performs multi-level feature extraction on the multi-modal fusion image data, including: texture scale decomposition, key point detection, and spectral difference measurement;

[0026] Define a deep representation mapping to map the pixel and neighborhood information of the multi-modal fusion image data into a high-dimensional feature vector containing multi-modal information representations of regions such as the scalp and hair follicles.

[0027] Preferably, the potential lesions or abnormalities are classified and detected through an anomaly recognition model, and a health score and analysis result data including basic detection results are output.

[0028] Combined with the environmental correction term and spatial distribution information, a cumulative measure is performed on the scalp area in a matrix-weighted manner, and finally a global health score is obtained.

[0029] Preferably, after generating the analysis result data, it is transmitted back to the corresponding module in the previous step, including:

[0030] If obvious local texture distortion or abnormal confidence is detected, the location information of the area and the corresponding environmental factors can be transmitted back to prompt to improve the lighting or maintain stability during the next shooting;

[0031] For areas with repeated segmentation errors, the analysis result data can be used as an increment of the training set or hard example annotation;

[0032] If it is found that a specific band contributes insufficiently to the detection, the band selection strategy or alignment accuracy can be dynamically adjusted during subsequent acquisitions.

[0033] Preferably, after obtaining the user's historical usage records, the existing deep learning segmentation model or recognition model is updated regularly or in real time based on the user's incremental data and environmental indicators.

[0034] Preferably, the updated model parameters are used to perform multi-label classification or segmentation on other scalp pathologies other than oiliness and inflammation, and after defining a multi-label discrimination function, it is output as a confidence vector of multiple categories:

[0035] If necessary, it can be extended to multi-label segmentation, that is, multiple pathological regions are simultaneously labeled on the basis of the mask of the aforementioned scalp area to achieve a finer-grained lesion map.

[0036] Preferably, the detection data from multiple users are subjected to population statistics or clustering analysis.

[0037] Define a global distribution function to characterize the correlation between different pathological labels and environmental factors. For each user, extract the corresponding pathological label gradient intensity in each environmental factor dimension from the global distribution function, and obtain a standardized feature vector after processing.

[0038] Select spectral clustering or Gaussian mixture model to cluster the standardized feature vectors. For each clustering cluster, assign a risk label according to the gradient intensity distribution of the cluster center in each pathological dimension.

[0039] According to the risk label, preferentially increase the infrared thermal imaging frame rate for high-inflammation-sensitive populations, automatically enable ultraviolet spectral band acquisition for moisture-prone fungal populations; and recommend corresponding treatment strategies for high-inflammation-sensitive populations and moisture-prone fungal populations.

[0040] An intelligent scalp detection system, including a data acquisition unit. When the detected light intensity and distance meet the available threshold, the environmental perception module combines with the image quality detection unit to calculate the mapped environmental score in real time and output the environmental and preliminary quality data.

[0041] An image processing unit. When the environmental and preliminary quality data indicate that imaging is available and the mapped environmental score is higher than the corresponding threshold, the deep learning segmentation model performs hair and scalp separation on the captured image and removes the noise area, calls the PDE-based enhancement component to combine denoising and super-resolution correction, and outputs the segmented and enhanced image data.

[0042] A data fusion unit. If the segmented and enhanced image data stably presents scalp features, the multimodal fusion module reads the visible light and near-infrared images, matches the overlapping areas through optimal transformation and performs local compensation to generate multimodal fusion image data.

[0043] A data analysis unit. The intelligent analysis model calls the global health score based on the environmental and preliminary quality data and the segmentation result to judge pathological abnormalities and outputs the analysis result data. At the same time, it transmits back the low-confidence regions for the previous segmentation and environmental adaptive optimization iteration.

[0044] A push unit. When a large amount of analysis result data is generated during the long-term use of the user and the pathological recognition is extended, combined with multi-label detection to update the model parameters by federated learning, and use the global distribution function to reveal the coupling between fungal or sebaceous gland abnormalities and environmental factors at the population dimension, and push the corresponding treatment strategies to the user.

[0045] (III) Beneficial effects

[0046] The present invention provides an intelligent scalp detection system and its detection method, which have the following beneficial effects:

[0047] Environment and Preliminary Quality Data (EPQD) and Mapped Environment Score Parameters such as can adaptively optimize lighting and focus at the shooting source, reducing image distortion caused by poor environment; can effectively suppress hair interference and highlight scalp details;

[0048] The segmented and enhanced image data is further matched with near-infrared or multi-spectral images. The generated multi-modal fusion image data can overcome the limitations of a single-band perspective. By performing local compensation in areas with severe occlusion or complex lighting, the surface and deep structures of the scalp are presented more completely and clearly. This fusion not only utilizes the shooting angle information but also combines the segmentation mask to perform adaptive matching and interpolation on overlapping or mismatched areas, integrating the advantages of multi-source images and greatly improving the accuracy of downstream recognition.

[0049] The intelligent analysis model uses multi-modal fusion image data to identify various scalp features such as sebum and inflammation, and adaptively tunes the model parameters based on the uncertainty of environmental quality and segmentation results. The resulting analysis result data can provide real-time diagnosis and be transmitted back to the previous steps. This multi-directional feedback mechanism ensures that the system can continuously optimize the segmentation and fusion quality in complex scenarios, reducing the probability of false alarms and missed detections.

[0050] During long-term use, the images collected from users multiple times and the corresponding pathological judgments are used to continuously update the model through online training or federated learning, incorporating multi-label detection, and then achieving accurate recognition of complex problems such as fungal infections and sebaceous gland abnormalities in a wider population. It also reveals the potential interaction laws between pathology and environmental factors. The combination of data from different bands, environmental feedback, and the results of deep models produces a synergistic effect, far exceeding traditional single-band or fixed algorithms in terms of scalp detection accuracy and stability, being able to meet the personalized needs of diverse users and scenarios and continuously iterate, providing a forward-looking and complete systematic solution for intelligent scalp health management. Brief Description of the Drawings

[0051] Figure 1 It is a schematic flow chart of the intelligent scalp detection method of the present invention;

[0052] Figure 2 It is a schematic structural diagram of the intelligent scalp detection system of the present invention. Detailed Embodiments

[0053] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0054] Please refer toFigure 1 , the present invention provides an intelligent scalp detection method, including,

[0055] Step 1: When initially collecting images, the environmental perception module senses the light intensity, irradiation angle, and camera distance, and calls the image quality detection unit to evaluate the clarity and scalp visibility, obtaining environmental and preliminary quality data and mapping environmental scores ;

[0056] The said Step 1 includes the following contents:

[0057] Step 101: Environmental parameter collection and comprehensive scoring

[0058] The environmental perception module respectively collects the light intensity, the distance between the camera and the scalp, and the movement parameters during user combing, and calculates the mapping environmental score according to the irradiation angle information , and encapsulates it into the environmental and preliminary quality data; where:

[0059] The main collection quantities of the environmental perception module include the light intensity; the distance value between the camera and the scalp surface, the inclination angle value of the camera relative to the scalp normal direction, and the movement smoothness;

[0060] To comprehensively reflect the usability of the shooting environment, the mapping environmental score is defined :

[0061]

[0062] Where:

[0063]

[0064]

[0065] The 2×2 matrix W to be calibrated can be calibrated through experimental data or set based on prior experience; is the index vector;

[0066] L is the light intensity, d is the distance; is the inclination angle of the camera relative to the scalp normal direction; m is the movement smoothness, is the desired smoothness;

[0067] Logistic function: , will from map to interval;

[0068] is the angle correction term, is the adjustable parameter used to control the influence degree of the angle on the shooting quality;

[0069] is the motion jitter penalty term, is the adjustment coefficient, ;

[0070] When in use, it can automatically identify situations of adverse environments (such as too dark light, too large distance, or severe jitter), and reduce the score in a timely manner, so that in subsequent steps, after receiving this score, it can decide whether to prompt the user to correct the shooting method;

[0071] Step 102, Preliminary Image Quality Detection and User Prompt

[0072] The image quality detection unit, based on the received mapped environment score and the initial captured image, calculates the image quality score using advanced methods such as gradient distribution or wavelet features This score will be encapsulated together with the mapped environment score into the environment and preliminary quality data. Among them, the environment and preliminary quality data EPQD contains environmental illumination, focal length, motion smoothness, image sharpness, and the mapped environment score etc.; and finally transmitted to Step Two, specifically as follows:

[0073] Using the mapped environment score and the initial captured image as the input of the image quality detection unit, in the following manner:

[0074]

[0075] In the formula: is the initial captured image;

[0076] represents the coefficient vector obtained after performing multi-scale wavelet transform on the initial captured image at scale s and direction o: Here, is the position of the image in the corresponding wavelet domain or the corresponding spatial domain;

[0077] represents the vector formed by the local gradient (or partial derivative) of at the same scale s and direction o, which is used to characterize the intensity and change trend of image details;

[0078] is a pre-defined or trained real symmetric weighted matrix (with dimensions matching ), and different s and o can correspond to different weight matrices to highlight or suppress specific frequency bands and direction information;

[0079] S and O respectively represent the scale set and direction set in the wavelet transform. For example , ; is the image coordinate domain, which can be the mapping area of wavelet coefficients in the spatial or frequency domain;

[0080] is the mapping environment score is the adjustment coefficient at time, and its value can be between 0.1 and 1.0, mainly controlling the magnification or suppression degree of the mapping environment score on the final result of the image quality score;

[0081] When the quality score and the mapping environment score both are lower than the preset threshold, the user prompt mechanism is triggered to recommend that the user adjust the shooting angle or slow down the moving speed. When both are relatively high or showing an upward trend, the subsequent steps can be continued without prompting;

[0082] When in use, on the premise of obtaining the mapping environment score perform multi-scale wavelet analysis on the image quality, which can achieve more stable image sharpness judgment, and at the same time achieve flexible control of different environmental states through logarithmic mapping;

[0083] The quality score and the mapping environment score jointly output the environment and preliminary quality data, fundamentally reducing the misjudgment risk of subsequent steps for poor images. This synergistic effect lies in real-time capturing and quantifying external interference factors, and can also use image features to mutually verify the actual imaging effect in both directions; and it realizes the non-linear characterization and adaptive optimization in complex shooting scenarios, providing a more reliable initial input for subsequent segmentation and multi-modal fusion.

[0084] Step 2: When the environment and preliminary quality data indicate that both the illumination and sharpness reach the available thresholds, the deep learning segmentation model performs hair-scalp pixel segmentation on the captured image, removes the interference areas, and adaptively adjusts the regularization coefficient in combination with the mapping environment score and performs denoising and enhancing scalp details to suppress hair occlusion, generating the segmented and enhanced image data;

[0085] The said Step 2 includes the following contents:

[0086] Step 201: Hair-scalp segmentation

[0087] Using the initial captured image as the image input, a pre-trained deep learning segmentation network is used to distinguish the pixels of hair and scalp, obtaining the probability that each pixel belongs to the scalp category; combining the pixel-level segmentation result of deep learning with the environmental quality feedback to automatically enhance the reliability of the segmentation decision in a poor environment or blurred area, where:

[0088] Define the following objective functional and obtain the segmentation mask after solving , where the segmentation mask is 1 for the scalp region and 0 for non-scalp regions (including hair and background), and the determination of the scalp region is achieved by the segmentation mask :

[0089]

[0090] In the formula: represents the negative log-likelihood term constructed based on the pixel label probabilities output by the deep learning segmentation network for the initial captured image , which is used to measure the consistency between the segmentation and the model prediction;

[0091] is a regularization term that combines the environment and the preliminary quality data EPQD. It mainly adaptively smooths or corrects the segmentation edges when the mapped environment score is low (such as poor lighting or obvious jitter); specifically as follows:

[0092] The regularization term is specifically defined as environment-adaptive weighted total variation smoothing, in the following form:

[0093]

[0094] where: is the value of the segmentation mask at pixel (1 for the scalp area and 0 for non-scalp);

[0095] is the image coordinate domain to be segmented; represents the gradient vector of the mask in the spatial dimension, which is used to measure the change rate of the mask boundary; is a small constant to prevent the denominator from being zero, such as ; is a weight parameter, and its value range depends on the experimental tuning results. It is used to balance the importance of the segmentation network signal and the environment correction information in the segmentation process;

[0096] When in use, the mapped environment score in the environment and the preliminary quality data EPQD is coupled into the segmentation process through the weight parameter . This can moderately smooth or correct the network decision in fuzzy or low-light scenarios, thereby reducing hair and scalp boundary errors; when the environment and the preliminary quality data EPQD score is good, the weight parameter Its role will be relatively weakened, mainly relying on network fine prediction; when the score is poor, the regularization effect will be relatively strengthened to improve segmentation stability. By reading information such as illumination and jitter in the reading environment and the preliminary quality data EPQD, the segmentation decision is corrected directionally, which helps to achieve higher accuracy in real complex scenes.

[0097] Step 202, Image enhancement

[0098] Focus on enhancing the scalp area determined by the segmentation mask to avoid artifacts caused by overprocessing hair pixels. Introduce an orientation enhancement model that combines partial differential equations (PDEs) and non-linear response functions to output the segmented and enhanced image data , and define the evolution equation as follows:

[0099]

[0100] In the formula: represents the enhanced image at the evolution time t. At the initial time , it can be set as ; is the divergence operator, is the gradient operator;

[0101] is the non-linear response function:

[0102] When (scalp area) and the EPQD indicates low illumination, this function takes a larger value to enhance detail repair; when (hair or background) or good illumination, this function takes a smaller value to reduce over-smoothing or enhancement of non-target areas;

[0103] , are constant coefficients, both with values greater than 0, used to control the diffusion strength and fidelity compensation term in the PDE;

[0104] This term ensures that the enhanced result retains the style of the original captured image while compensating for illumination, and will not be distorted into over-sharpening or over-smoothing;

[0105] When in use, the hair pixels are masked by the segmentation mask , and only the scalp area is refined to ensure a one-to-one correspondence between the segmented area and the enhancement requirements, avoiding unnecessary artifacts. The non-linear response function combines the quality score in the environment and the preliminary quality data EPQD , stronger local enhancement can be applied to areas with uneven illumination or obvious noise in the scalp area, which can improve the visualization effect and recognition accuracy. Different from common global filtering or global super-resolution, only the scalp mask area is weighted through the evolution of the environment and preliminary quality data EPQD, and environmental factors are incorporated, which has higher pertinence and controllability. After completing the above two steps, not only the confusion of hair occlusion for the details of the scalp area is effectively eliminated, but also the distinguishability of the scalp texture is improved by targeted enhancement.

[0106] Step 3: When the image data after segmentation and enhancement stably presents the scalp area, the multi-modal fusion module reads the visible light image and near-infrared data, combines the shooting angle and mask information, performs weighted stitching after affine mapping, and applies adaptive weights to multiple scalp lesions such as sebum and inflammation in the overlapping area for real-time recognition, outputs the analysis result data, and feeds back the low-confidence area to the segmentation model and shooting module for adaptive correction;

[0107] The above step 3 includes the following contents:

[0108] Step 301: Multi-modal data alignment

[0109] Collect multi-modal additional image data , which can include shooting in the near-infrared band, multi-spectral channels or other angles;

[0110] Align the image data after segmentation and enhancement with the multi-modal additional image data precisely in the same spatial coordinate system. To take into account the complex texture of the scalp and possible non-linear distortions, construct the following regional matching energy functional to solve the optimal transformation :

[0111]

[0112] represents the geometric transformation for mapping the multi-modal additional image data in the target coordinate domain, which can be composed of function families such as affine, bilinear or more advanced thin plate spline (TPS);

[0113] is the alignment consistency measure, which can be based on the Kullback-Leibler divergence or other high-dimensional distribution difference measurement methods of the segmentation area indicated by the scalp mask rather than simple pixel differences;

[0114] is the regularization term, which is used to limit the drastic distortion of the transformation when the environment is unfavorable (such as large jitter, weak illumination);

[0115] is the balance coefficient, and its value range is between 0.1 and 5.0;

[0116] Apply the optimal transformation to the multi-modal additional image data After that, the segmented and enhanced image data and the aligned multi-modal image data in the same coordinate system ;

[0117] When the environment and the preliminary quality data EPQD indicate high jitter or low light, the balance coefficient will be increased , so that the smoothness of the geometric transformation is enhanced, and the registration failure caused by local extreme changes is minimized; when the environment is better, more reliance is placed on the alignment consistency measure for high-precision matching of details;

[0118] When in use, the non-linear transformation is combined with the distribution information of the pixels in the segmentation mask, so that higher resolution and finer-level alignment can be achieved for multi-modal images in the scalp area: through the regularization term introduce the feedback of the environment and the preliminary quality data EPQD, and automatically adjust the transformation smoothness according to the shooting environment state to ensure the alignment reliability.

[0119] Step 302, Fusion Compensation and Stitching

[0120] The segmented and enhanced image data and the multi-modal additional image data These two images are weighted and superimposed in the same coordinate system with adaptive weights to generate multi-modal fusion image data ; where:

[0121] To balance the low-frequency consistency (global brightness and contrast) and high-frequency detail complementarity (hair follicle texture and shallow / deep structure), the following weighted fusion model is adopted in the scalp area in this step:

[0122] Let represent the multi-modal fusion image data at the evolution time t ( is the coordinate on the image plane, such as ), and at the initial moment, it can be set:

[0123]

[0124] That is, first use the visible light image as the initial solution, and define the following evolution equation:

[0125]

[0126] Where: is the divergence operator, is the gradient operator;

[0127] is the image data after segmentation and enhancement in the visible light domain; is the multi-modal additional image data obtained after alignment (such as near-infrared, multi-spectral, etc.);

[0128] is the scalp region mask weight function, and the value range is ;

[0129] is the diffusion tensor (or matrix) in the evolution equation, which is used to control how the image is smoothed and gradient diffused during the iteration process. It can adaptively adjust the diffusion intensity of certain regions according to the mapped environment score data (EPQD); is the data fidelity coefficient, which also combines EPQD information and the value can be between 0.01 and 2.0;

[0130] Among them: Initialization: Set as ;

[0131] Iterative evolution: Update sequentially in the discrete time domain until convergence or reaching the preset iteration upper limit;

[0132] Convergence result: When approaches 0, a stable multi-modal fusion image data is obtained, which contains the matching results of visible light and near-infrared / multi-spectral bands;

[0133] For the iterative solution of the above PDE-based multi-modal fusion formula, if you want to accelerate convergence, you can add parameters such as illumination and distance during initialization or iteration to the diffusion tensor and the data fidelity coefficient for dynamic adjustment, so as to approach the ideal fusion state faster;

[0134] When in use, by referring to the alignment residual and the mapped environment score , local anomalies can be processed directionally, minimizing stitching artifacts and artifacts. Introducing alignment residuals and EPQD information during the fusion process can achieve precise suppression or repair of mismatched regions, significantly improving the final stitching quality.

[0135] In Step 3, multi-modal fusion image data is finally output. The fusion result presents comprehensive information of multiple bands or multiple perspectives in the scalp area, overcoming the occlusion and brightness limitations that may be encountered in pure visible light images, and also using additional information such as near-infrared or multi-spectral to reveal deeper structural features of hair follicles and the scalp. Combining the EPQD, segmentation masks, and illumination enhancement information accumulated in the previous steps, it demonstrates strong adaptability and reliability in the multi-modal data stitching and compensation process.

[0136] Step 4: When the multi-modal fusion image data is stitched and has sufficient clarity, the intelligent analysis model imports the environmental and preliminary quality data and the segmentation results, calls the anomaly recognition model to perform real-time recognition of various scalp lesions such as sebum and inflammation, and outputs the analysis result data. At the same time, the low-confidence regions are sent back to the segmentation model and the shooting module for adaptive correction;

[0137] The content of Step 4 is as follows:

[0138] Step 401: Feature parsing and deep representation

[0139] The intelligent analysis model performs multi-level feature extraction on the multi-modal fusion image data including but not limited to:

[0140] Texture scale decomposition: Use multi-scale or multi-directional filters (such as Gabor filtering, Curvelet, etc.) to perform hierarchical detection on the scalp texture from coarse to fine;

[0141] Key point detection: Extract local feature vectors of hair follicle openings, inflammation hotspots, or other abnormal points;

[0142] Spectral difference measurement: For multi-band information, calculate the dispersion of the feature distribution within the scalp area (such as using a local energy function or a similarity measure based on spectral curve fitting) to determine abnormal sebum secretion or pigment deposition, etc.;

[0143] Define the following deep representation mapping , and map the pixel and neighborhood information of the multi-modal fusion image data to a high-dimensional feature vector containing multi-modal information representation of regions such as the scalp and hair follicles :

[0144]

[0145] In the formula: represents the image coordinates; is the local gradient information, used to identify edges or rapidly changing regions;

[0146] can refer to the pixel response vector under the multi-spectral channel; Unify and encapsulate the spectral, texture, and depth features of the scalp surface, laying a foundation for subsequent classification or clustering;

[0147] During use, integrate visible light, near-infrared / multispectral, and enhanced scalp region information in the same feature map, which can more comprehensively reflect scalp health indicators and achieve multi-modal integrated analysis; fuse gradient / texture features of multiple bands and segmented regions into a unified high-dimensional representation, rather than simple splicing or sequential processing, which can significantly improve the discriminability of complex scalp features and perform deep representation mapping Perform multi-scale and multi-channel depth encoding on scalp region features, reducing missed detections or false detections caused by a single scale or single band.

[0148] Step 402, Abnormality recognition and health scoring

[0149] After obtaining the high-dimensional feature vector classify and detect potential lesions or abnormalities through an abnormality recognition model, and output health scores and analysis result data , including basic detection results such as sebum secretion, inflammation, and hair follicle health;

[0150] Among them, a deep neural network (such as Transformer or self-attention mechanism) or a hybrid model (such as the combination of random forest and graph neural network) can be used to learn the high-dimensional feature vector ; the abnormality recognition model is obtained by training a convolutional neural network (such as ResNet, EfficientNet) or a vision Transformer (ViT) architecture, and adding an attention mechanism or a graph neural network layer;

[0151] Combined with environmental correction terms and spatial distribution information, perform cumulative measurement on it in the scalp region through matrix weighting, and finally obtain a global health score :

[0152]

[0153] In the formula: is a deep discriminant function, which is a neural network model. It can typically be represented as a multi-layer perceptron (MLP) or a lightweight convolutional network. The last layer uses Sigmoid (binary classification) or Softmax (multi-classification) activation to output the probability value that the pixel or local belongs to the abnormal (such as inflammation, excessive sebum) or healthy category, or perform other forms of risk measurement;

[0154] is the high-dimensional feature vector of the multi-modal fusion image at pixel ; Denote the weights of neural networks or algorithms for recognition / classification tasks, which have been determined or dynamically updated during the training phase;

[0155] Denote the discriminant function For the coordinates in the feature space Gradient vector;

[0156] Is a real symmetric weighted matrix used to adjust the contribution of gradients at different positions to the final integration result, and its size and direction can be dynamically changed according to the environment correction function Dynamically change:

[0157] Is a real symmetric positive definite matrix (such as 2×2 or n×n) matching the dimension of the gradient vector, which weights the contributions of gradient components in different directions;

[0158] Is the environment correction function, which is linked to the previously defined EPQD (Environment and Preliminary Quality Data), and converts the environment and preliminary quality data EPQD into confidence weights for each pixel In subsequent processing;

[0159] Is an additional global correction or compensation term for overall adjustment outside the integral term; if no global correction or compensation is required, Can be regarded as a constant or set to zero.

[0160] When in use, the environment correction function Makes the model cautious in areas with poor lighting and reduces the risk of false alarms: in high-quality areas, it gives full play to the detection ability to achieve environment-adaptive anomaly recognition; by non-linearly integrating the classification results of all pixels / blocks within the scalp area, a highly integrated global health score can be output To help users understand the overall health status; coupling environmental information such as lighting and jitter with the deep discriminant model in the same integral functional, an environment-adaptive health score formula is constructed, which is more flexible and stable than traditional single classification or simple weighting.

[0161] Step 403, Closed-loop feedback and adaptive model update

[0162] After generating the analysis result data, these results are all sent back to the corresponding modules in the previous steps to form a closed loop. Typical feedback paths include but are not limited to:

[0163] Sent back to the environmental perception module in step one:

[0164] If significant distortion of local texture or abnormal confidence is detected in step 402, the location information of the area and the corresponding environmental factors (such as lighting or jitter in EPQD) can be transmitted back to prompt increasing lighting or maintaining stability during the next shooting;

[0165] Transmit back to the deep learning segmentation model in step two:

[0166] For areas with repeated segmentation errors, the analysis result data can be used as an increment to the training set or hard example annotation to further improve the segmentation accuracy of the model under fine textures and complex hair qualities;

[0167] Transmit back to the multi-modal fusion module:

[0168] If step 402 finds that a specific band contributes insufficiently to the detection, the band selection strategy can be dynamically adjusted in subsequent acquisitions, or the alignment accuracy can be improved;

[0169] During use, by continuously flowing the analysis result data back to the previous stage, its own defects can be continuously corrected and it can continuously iterate and evolve: if the user's operating habits cause a lot of jitter or the lighting environment is complex, the segmentation model or fusion strategy can also be adaptively adjusted after closed-loop feedback to reduce inaccurate recognition caused by human errors, thereby lowering the user's operation threshold. During subsequent shootings, certain specific bands can be used preferentially according to the instructions of the closed-loop feedback, or higher-resolution re-acquisition can be performed in some key areas.

[0170] Step five: When the user accumulates a large amount of analysis result data during long-term use, the online training module and the multi-label detection module extract new pathological dimensions such as fungal infections or sebaceous gland abnormalities, and iteratively update the model parameters through federated learning and fuse environmental factors in to accurately identify more pathological types and push corresponding treatment strategies to the user;

[0171] The said step five includes the following contents:

[0172] Step 501, Long-term data collection and online training

[0173] Obtain the user's historical usage records: including the user's multiple detection results and possible subjective feedback (such as whether the user marks that there is real inflammation), and this part of the data accumulates continuously during long-term use;

[0174] Regularly or real-time update the existing deep learning segmentation model or recognition model through the online training module;

[0175] To be compatible with the scenarios of federated learning and local online learning, define the iterative update formula. Let represent the model parameters at time t, Represents newly collected batch data (which contains user feedback or high-confidence samples after automatic screening), you can construct:

[0176]

[0177] in: is a gradient function or differential mapping, used according to Correct the current model based on segmentation errors, classification errors or environmental factors; is the learning rate or update step size, which can be dynamically adjusted by the illumination and jitter indicators in EPQD;

[0178] During use, the model can gradually adapt to individual characteristics. During multiple tests, the model will adapt to the user's specific scalp environment, greatly reducing the long-term detection error of the same user. When new symptoms or extreme lighting conditions are detected, new batches of data The difficult examples in will improve the reliability of the model during iterative updates.

[0179] Step 502: Multi-label pathology recognition expansion

[0180] Perform multi-label classification or segmentation on scalp pathologies other than oil and inflammation (such as fungal infection, sebaceous gland dysfunction, etc.), and define a multi-label discriminant function , whose output is the confidence vector of multiple categories:

[0181]

[0182] Where: Each component in corresponds to a pathological label (such as fungal sebaceous gland abnormal inflammation secondary infection, etc.), It is a high-dimensional feature vector, or new features are added after online training;

[0183] With the updated model parameters , more detailed pathological information can be identified;

[0184] If necessary, it can be extended to multi-label segmentation, that is, multiple pathological areas are labeled simultaneously based on the mask of the aforementioned scalp area to achieve a more fine-grained lesion map.

[0185] When in use, the coverage of disease types is expanded, and more possible pathological types are added on the basis of the original oil, inflammation, etc., which can be closer to clinical needs, and can still output traditional indicators for oil or inflammation without destroying the core results of the original recognition process; it can simultaneously output the confidence or segmentation masks of multiple pathologies, breaking through the limitations of traditional binary classification or single label. If the newly emerging pathological types are less in the early samples, after accumulating more real samples, the recognition performance of the new labels can be rapidly improved.

[0186] Step 503: Cross-user big data statistics and group feature modeling

[0187] Perform group statistics or clustering analysis on the detection data from multiple users to mine more representative scalp health patterns and define the global distribution function Characterize the association between different pathological labels and environmental factors of, Represent the optional pathology-environment attention vector (such as the parameters corresponding to specific pathological labels or composite indicators), where:

[0188]

[0189] Represents the index domain or measure space of the user set, that is, there are multiple users u in the federated learning or large-scale application scenario; Represents the environmental factor domain, which can cover the light intensity, humidity, region or other extended scenario variable integrals;

[0190] Is the aggregated analysis result data, which is usually associated with the user u and includes the pathological test results (such as labels or probabilities of inflammation, sebum, fungal infection, etc.) and the aggregated features of the scalp images;

[0191] Is the aggregation function used to couple and map the pathological information of the user with the environmental factors : It can internally calculate the complex association between the pathological label and the environmental variable, such as the sebum secretion and the indoor temperature, the fungal infection and the humidity, etc.;

[0192] Represents the gradient vector obtained by this mapping in the composite dimension;

[0193] Is a real symmetric weighted matrix (or tensor), which is closely associated with the parameter . Different parameters can specify to emphasize or suppress certain pathology-environment correlation dimensions, for example, pay special attention to the fungus + humidity dimension;

[0194] Is the global correction or compensation term outside this double integral, and its functions can include:

[0195] Overall weight increase for extreme scenarios (such as ultra-high humidity, rare pathological labels) to avoid being masked by sparse data;

[0196] Deduct or compensate for certain common factors (such as the unified inherent error of the device) once. If there is no such requirement, it can be regarded as a constant or set to 0.

[0197] Obtain the global distribution function After that, for each user , extract the corresponding pathological label gradient intensity in each environmental factor dimension from the global distribution function , splice them into a vector , perform column-wise normalization on all to eliminate the dimensional differences of different environmental factors, and obtain ; Select spectral clustering (SpectralClustering) or Gaussian mixture model (GMM) to cluster. For each cluster, according to the gradient intensity distribution of the cluster center in each pathological dimension, assign risk labels:

[0198] If a certain cluster has a high gradient in the inflammation-temperature or oil-moisture dimension, it is marked as highly inflammation-sensitive;

[0199] If a certain cluster has a high gradient in the fungus-moisture or sebaceous gland activity-moisture dimension, it is marked as prone to fungus in humid environment;

[0200] The remaining clusters are named according to the actual gradient characteristics (such as low risk / hair follicle weakening, etc.).

[0201] For the highly inflammation-sensitive population, increase the infrared thermal imaging frame rate preferentially, and automatically enable ultraviolet spectral band acquisition for the population prone to fungus in humid environment;

[0202] Recommend corresponding treatment strategies for the highly inflammation-sensitive population and the population prone to fungus in humid environment, for example: recommend anti-inflammatory formulas for the highly inflammation-sensitive population, and recommend antibacterial lotions for the population prone to fungus in humid environment.

[0203] When in use, it can evaluate the distribution of scalp problems in a specific region or specific population, help develop more targeted nursing programs, and through the group characteristics learned from cross-user data, make the model more capable of dealing with diverse scalp types and environmental variables. It can be linked with the online learning module to dynamically update the mapping mode according to newly emerged pathological types or special scenarios, and improve the accuracy of statistics and clustering.

[0204] Through online training or federated learning, the system can accumulate data of different users for a long time and continuously iterate model parameters, use the updated model to achieve multi-label detection, expand the coverage of the system for scalp pathology, perform group statistics on the detection results of multiple users and multiple scenarios, form a more general understanding of pathological distribution, and then feedback to steps 501 and 502 again, which can continuously optimize the model and detection range, while maintaining the accurate detection and analysis capabilities constructed in the first four steps, and having the potential for sustainable evolution facing long cycles, multiple scenarios, and large populations.

[0205] Please refer to Figure 2, the present invention provides an intelligent scalp detection system, including,

[0206] A data acquisition unit. When it detects that the light intensity and distance meet the available thresholds, the environmental perception module combines with the image quality detection unit to calculate the mapped environmental score in real time and output the environmental and preliminary quality data;

[0207] An image processing unit. When the environmental and preliminary quality data indicates that imaging is available and the mapped environmental score is higher than the corresponding threshold, the deep learning segmentation model performs hair and scalp separation on the captured image and removes the noise area, calls the PDE-based enhancement component to combine denoising and super-resolution correction, and outputs the segmented and enhanced image data;

[0208] A data fusion unit. If the segmented and enhanced image data stably presents scalp features, the multimodal fusion module reads the visible light and near-infrared images, matches the overlapping areas through optimal transformation and performs local compensation to generate multimodal fusion image data;

[0209] A data analysis unit. The intelligent analysis model calls the global health score to distinguish pathological abnormalities based on the environmental and preliminary quality data and the segmentation results, and outputs the analysis result data. At the same time, it transmits back the low-confidence regions for the previous segmentation and environmental adaptive optimization iteration;

[0210] A push unit. When the user generates a large amount of analysis result data during long-term use and expands pathological recognition, it combines multi-label detection to update the model parameters by federated learning, and uses the global distribution function to reveal the coupling between fungal or sebaceous gland abnormalities and environmental factors at the population level, and pushes the corresponding treatment strategies to the user.

[0211] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but this implementation should not be considered to exceed the scope of this application.

[0212] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0213] In several embodiments provided in the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only for some logical function divisions. In actual implementation, there may be other division methods. 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 displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0214] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0215] As described above, the above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed in the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. An intelligent scalp detection method, characterized in that: including, when it is detected that the light intensity and the distance value of the camera from the scalp surface satisfy the available threshold, the environmental perception module combines with the image quality detection unit to calculate the mapped environmental score in real time and output the environmental and preliminary quality data; wherein, the environmental perception module collects the light intensity, the distance value of the camera from the scalp surface, the tilt angle value of the camera relative to the scalp normal direction, and the motion smoothness, and thereby defines the mapped environmental score to reflect the usability of the shooting environment; wherein, the environmental and preliminary quality data includes the quality score, the mapped environmental score, the environmental light, the focal length, the motion smoothness, and the image sharpness. If the mapped environmental score is higher than the corresponding threshold, the deep learning segmentation model performs hair and scalp separation on the captured image, eliminates the noise area, calls the PDE-based enhancement combined with denoising and super-resolution correction, and outputs the segmented and enhanced image data. If the segmented and enhanced image data stably presents the scalp characteristics, the multimodal fusion module reads the visible light and near-infrared images, matches the overlapping areas through the optimal transformation and performs local compensation, and generates the multimodal fusion image data. The intelligent analysis model calls the global health score to discriminate pathological abnormalities based on the environmental and preliminary quality data and the segmentation results, and outputs the analysis result data. At the same time, it transmits the low-confidence areas back for the previous segmentation and environmental adaptive optimization iteration. Combined with multi-label detection, the model parameters are updated by federated learning, and the global distribution function is used to reveal the coupling between fungal or sebaceous gland abnormalities and environmental factors at the population dimension, and corresponding treatment strategies are pushed to the user.

2. The intelligent scalp detection method according to claim 1, wherein: The image quality detection unit calculates the image quality score based on the received mapped environmental score and the initial captured image by using the gradient distribution or wavelet features. When both the quality score and the mapped environmental score are lower than the corresponding preset thresholds, the user prompt mechanism is triggered.

3. The intelligent scalp detection method according to claim 2, wherein: Using the initial captured image as the image input, the pre-trained deep learning segmentation network differentiates the pixels of the hair and the scalp, combines the pixel-level segmentation results of the deep learning with the environmental quality feedback, defines the objective functional and obtains the segmentation mask after solving, identifies the scalp area and the non-scalp area according to the segmentation mask, and adaptively smooths or corrects the segmentation edge.

4. The intelligent scalp detection method according to claim 3, wherein: For the scalp area determined by the segmentation mask, a directional enhancement model combining partial differential equations and non-linear response functions is introduced to output the segmented and enhanced image data.

5. The intelligent scalp detection method according to claim 4, wherein: Collect multimodal additional image data, including near-infrared bands, multispectral channels, or shooting from other angles. Precisely align the segmented and enhanced image data and the multimodal additional image data in the same spatial coordinate system, construct the regional matching energy functional and solve for the optimal transformation. After applying the optimal transformation to the multimodal additional image data, the segmented and enhanced image data and the aligned multimodal image data in the same coordinate system are obtained.

6. The intelligent scalp detection method according to claim 5, wherein: The segmented and enhanced image data and the multi-modal additional image data are weighted and superimposed in the same coordinate system with adaptive weights to generate multi-modal fusion image data, including the matching results of visible light, near-infrared, and multi-spectral bands.

7. The intelligent scalp detection method according to claim 6, wherein: The intelligent analysis model performs multi-level feature extraction on the multi-modal fusion image data, including texture scale decomposition, key point detection, and spectral difference measurement; Define a deep representation mapping to map the pixel and neighborhood information of the multi-modal fusion image data into a high-dimensional feature vector containing the multi-modal information representation of the scalp and hair follicle regions.

8. The intelligent scalp detection method according to claim 7, wherein: The potential lesions or abnormalities are classified and detected through an anomaly recognition model, and a health score and analysis result data including basic detection results are output; Combined with the environmental correction term and spatial distribution information, a cumulative measure is performed on the scalp area in a matrix weighting manner, and finally a global health score is obtained.

9. The intelligent scalp detection method according to claim 8, wherein: After generating the analysis result data, it is transmitted back to the corresponding modules in the previous steps, including: If obvious distortion of local texture or abnormal confidence is detected, the corresponding region position information and corresponding environmental factors are transmitted back to prompt to increase the illumination or maintain stability during the next shooting; For regions with repeated segmentation errors, the analysis result data can be used as an increment of the training set or hard example annotation; If it is found that a specific band contributes insufficiently to the detection, the band selection strategy or alignment accuracy can be dynamically adjusted during subsequent acquisitions.

10. The intelligent scalp detection method according to claim 9, wherein: After obtaining the user's historical usage records, based on the user increment data and environmental indicators, the existing deep learning segmentation model or recognition model is updated regularly or in real time.

11. The intelligent scalp detection method according to claim 10, wherein: The updated model parameters are used to perform multi-label classification or segmentation on other scalp pathologies other than oiliness and inflammation, and after defining a multi-label discrimination function, it is output as a confidence vector of multiple categories; Based on the mask of the aforementioned scalp region, multiple pathological regions are simultaneously marked to generate a finer-grained lesion map.

12. The intelligent scalp detection method according to claim 11, wherein: The detection data from multiple users are subjected to population statistics or clustering analysis; Define a global distribution function to characterize the correlation between different pathological labels and environmental factors. For each user, extract the corresponding pathological label gradient intensity in each environmental factor dimension from the global distribution function, and obtain a standardized feature vector after processing; Select spectral clustering or Gaussian mixture model to cluster the standardized feature vectors. For each clustering cluster, according to the gradient intensity distribution of the cluster center in each pathological dimension, a risk label is assigned; According to the risk label, the infrared thermal imaging frame rate is preferentially increased for high-inflammation sensitive populations, the ultraviolet spectral band acquisition is automatically enabled for moisture-prone fungal populations, and corresponding treatment strategies are recommended for high-inflammation sensitive populations and moisture-prone fungal populations.

13. Intelligent scalp detection system, characterized in that: including a data acquisition unit, when it is detected that the light intensity and the distance value of the camera from the scalp surface satisfy the available threshold, the environmental perception module combines with the image quality detection unit to calculate and output the mapped environmental score and the environmental and preliminary quality data in real time; wherein, the environmental perception module collects the light intensity, the distance value of the camera from the scalp surface, the tilt angle value of the camera relative to the scalp normal direction, and the motion smoothness, and thereby defines the mapped environmental score to reflect the usability of the shooting environment; wherein, the environmental and preliminary quality data includes the quality score, the mapped environmental score, the environmental light, the focal length, the motion smoothness, and the image clarity an image processing unit, if the mapped environmental score is higher than the corresponding threshold, the deep learning segmentation model performs hair and scalp separation on the captured image and removes the noise area, calls PDE-based enhancement combined with denoising and super-resolution correction, and outputs the segmented and enhanced image data a data fusion unit, if the segmented and enhanced image data stably presents scalp features, the multimodal fusion module reads the visible light and near-infrared images, matches the overlapping areas through optimal transformation and performs local compensation to generate multimodal fusion image data a data analysis unit, the intelligent analysis model calls the global health score based on the environmental and preliminary quality data and the segmentation result to discriminate pathological abnormalities and outputs the analysis result data, and at the same time transmits back the low-confidence areas for the previous segmentation and environmental adaptive optimization iteration a push unit, combines multi-label detection to update the model parameters by federated learning, and uses the global distribution function to reveal the coupling of fungal or sebaceous gland abnormalities and environmental factors at the population dimension, and pushes the corresponding treatment strategies to the user

Images