Skin moisture analysis system based on computer vision technology

Abstract

This invention relates to the field of image analysis technology, specifically a skin moisture analysis system based on computer vision technology. The system includes a skin region segmentation module, a moisture distribution modeling module, a texture enhancement and optimization module, and a moisture distribution classification module. In this invention, by accurately analyzing the boundaries, area, pixel values, and brightness values ​​of skin image regions, the moisture distribution state is defined and a weighted structure map is generated, reflecting the moisture differences between regions and providing accurate data support for subsequent analysis. In-depth analysis of node relationships and moisture value differences reveals moisture aggregation patterns and forms a highly accurate distribution law model, precisely displaying moisture characteristics. By extracting gradient values ​​from densely moist areas and utilizing texture contrast enhancement technology, moisture differences are made more significant, optimizing image details and presenting a more delicate moisture distribution map. Through hierarchical classification, the moisture state is accurately reflected, supporting differentiated analysis and significantly improving analytical accuracy and application value.

Description

Technical Field

[0001] This invention relates to the field of image analysis technology, and in particular to a skin moisture analysis system based on computer vision technology. Background Technology

[0002] Image analysis technology is a comprehensive discipline that combines computer vision, image processing, and artificial intelligence algorithms. It is mainly used to extract meaningful information from image or video data. Its core includes image acquisition, preprocessing, feature extraction, pattern recognition, and presentation of analysis results. The main methods involved in this field include edge detection, segmentation algorithms, texture analysis, deep learning models (such as convolutional neural networks), and classifier design. It is widely used in scenarios such as medical image diagnosis, industrial inspection, behavior monitoring, biometric recognition, and environmental monitoring. Its key goal is to identify, classify, quantify, or 3D reconstruct target characteristics or regions in images through efficient algorithms to support decision-making in scientific research or practical applications.

[0003] Among them, the skin moisture analysis system is an application tool based on image analysis technology, specifically used to assess the moisture content and distribution of the skin surface and its structure. The system collects images of skin areas and combines image processing and feature analysis technology to calculate skin moisture-related parameters, thereby achieving quantitative assessment and visualization of skin moisture status. Its main uses include verifying the effectiveness of skin care products, developing personalized skin care plans, and assisting in the diagnosis of skin diseases in clinical medicine, providing technical support for skin science research and daily care.

[0004] Current technologies for processing skin moisture distribution rely on traditional image processing methods, such as simple comparison of color and brightness values. These methods fail to accurately capture subtle differences in moisture levels across different areas of the skin. Most existing systems lack sufficient precision in boundary recognition and moisture distribution analysis, resulting in an overly blurry overall distribution of skin moisture that fails to reflect subtle changes. Furthermore, the lack of adequate weighting and optimization in region segmentation and moisture modeling results in inaccurate moisture prediction capabilities. Existing systems also offer relatively simple texture optimization, which is insufficient to effectively enhance the detail of skin images. Consequently, they cannot realistically reproduce the skin's moisture state and texture. In practical applications, some skin areas exhibit excessive moisture concentration, and traditional methods cannot distinguish subtle regional differences, leading to decreased accuracy in the analysis results. These shortcomings not only affect the accuracy of verifying the effectiveness of skincare products but also restrict the development of personalized care plans and the accuracy of medical diagnoses, ultimately limiting the practical application value of image analysis technology in skin moisture assessment. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and propose a skin moisture analysis system based on computer vision technology.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a skin moisture analysis system based on computer vision technology includes:

[0007] The skin region segmentation module extracts the color and brightness values ​​of pixels based on the input skin image, analyzes the location and area of ​​region boundaries, summarizes the region adjacency relationship, establishes the region segmentation result, labels the weight of the water content range of region nodes, performs weighted analysis on the node connection relationship, and generates a weighted structure of skin regions.

[0008] The moisture distribution modeling module extracts the edge weight distribution values ​​between nodes based on the weighted structure of the skin region. By superimposing the characteristic difference values ​​and the global characteristic values, it performs moisture clustering operations according to the distribution pattern to obtain moisture distribution pattern parameters. It compares the differences in moisture values ​​between regions and analyzes the corresponding distribution relationship between regions through the difference values ​​to obtain a moisture distribution pattern model.

[0009] The texture enhancement and optimization module extracts grayscale gradient values ​​in areas with dense moisture based on the moisture distribution pattern model, adjusts grayscale values ​​according to texture gradient characteristics, optimizes pixel differences within the area, performs texture contrast enhancement, refines the enhanced texture features in layers, generates a skin texture enhancement structure, and performs global matching of regional characteristic values ​​to obtain a texture optimization feature map.

[0010] Based on the texture optimization feature map, the moisture distribution classification module extracts the local area moisture content value and global distribution characteristics, classifies the area moisture content range into levels, matches the distribution pattern features, compares the distribution levels by region, and generates skin moisture distribution classification results.

[0011] As a further aspect of the present invention, the step of obtaining the region division result specifically includes:

[0012] Based on the input skin image, the color value and brightness value of each pixel in the skin image are extracted. A color distribution matrix is ​​constructed for the color value of the pixel, and a brightness distribution matrix is ​​constructed for the brightness value. The overall distribution range of color and brightness values ​​is analyzed based on the color distribution matrix and the brightness distribution matrix to obtain the preliminary distribution range parameters of color and brightness.

[0013] Based on the preliminary color and brightness distribution range parameters, difference analysis is performed on the color and brightness values ​​of adjacent pixels. A color and brightness difference threshold is set based on the calculation results. Boundary analysis is then performed on adjacent pixel pairs using the following formula:

[0014]

[0015] Calculate the boundary values ​​of pixels, mark pixels with values ​​greater than the threshold as boundary points, and generate a region boundary matrix;

[0016] Among them, B i,j C represents the boundary value of a pixel. i,j and C i+1,j L represents the color values ​​of adjacent pixels respectively. i,j and L i,j+1 W represents the brightness value of adjacent pixels respectively. c and W l These are the normalized weighting coefficients for color and brightness, respectively.

[0017] Based on the region boundary matrix, the regions are identified and segmented by performing connectivity analysis on the marked boundary points, the area of ​​the segmented regions is counted, the shape parameters of each region are calculated based on the segmentation results, the adjacency relationships between regions are analyzed, and the region division results are established.

[0018] As a further aspect of the present invention, the step of obtaining the weighted structure of the skin region specifically includes:

[0019] Based on the region division results, the range of regional nodes is extracted, and the moisture content data of the nodes in the region is extracted. The upper and lower limits are analyzed, abnormal data is filtered, and the moisture data of edge nodes is corrected to obtain a table of the range of regional node moisture content.

[0020] Based on the regional node moisture content range table, the upper and lower limits of node moisture content are normalized, the normalization weights are adjusted, the weight balance is optimized, the weight values ​​are recalibrated, and the regional node weighted table is obtained.

[0021] Based on the weighted table of regional nodes, the node connection relationship is analyzed, the geometric distance and normalized weight difference of moisture content between nodes are calculated, the weight matrix is ​​corrected, the node information and the weight matrix are integrated, and a weighted structure of the skin region is generated.

[0022] As a further aspect of the present invention, the steps for obtaining the moisture distribution pattern parameters are specifically as follows:

[0023] Based on the weighted structure of the skin region, the edge weight values ​​between each node are extracted. By calculating the local characteristic difference values ​​between nodes and superimposing the global characteristic values, the edge weight values ​​are initially summarized. Combining the distribution characteristics of the edge weight values ​​between nodes with the summary results, the weighted node distribution matrix is ​​obtained.

[0024] Using the weight parameters recorded in the weighted node distribution matrix, a quantitative analysis of the characteristic difference value and the global characteristic value is performed, using the formula:

[0025]

[0026] Obtain the characteristic distribution difference matrix;

[0027] Among them, T kW represents the difference in characteristic distribution. a and W b D represents the local and global weight values, respectively. k For local characteristic values, G a is the global characteristic value, and n and m are the total number of local and global characteristics, respectively;

[0028] Based on the aforementioned characteristic distribution difference matrix, clustering operations are performed on the nodes by analyzing the weighted weights between the nodes, and the water area characteristics are divided according to the node clustering pattern. The relationship between the node distribution pattern and the regional characteristics is extracted to obtain the water distribution pattern parameters.

[0029] As a further aspect of the present invention, the steps for obtaining the moisture distribution pattern model are as follows:

[0030] Based on the aforementioned moisture distribution pattern parameters, the moisture content values ​​of nodes within the region are extracted, the node data are grouped by region, the moisture value distribution pattern is fitted, and a regional moisture distribution statistical table is generated.

[0031] Based on the aforementioned regional moisture distribution statistics table, the statistical parameters of regional moisture distribution are compared, the deviations in moisture distribution patterns between regions are analyzed, the correlation patterns between regions are integrated, and the data on regional moisture distribution differences are obtained.

[0032] Based on the data on regional water distribution differences, the differences in water distribution between regions and the correlation parameters are extracted, the parameters are aggregated into a unified model, the regional differences are mapped to parameter curves, the water distribution parameters are integrated, and a water distribution pattern model is obtained.

[0033] As a further aspect of the present invention, the step of obtaining the skin texture enhancement structure specifically includes:

[0034] Based on the moisture distribution pattern model, the pixel gray values ​​in the densely moist areas are extracted, the gray gradient values ​​are calculated, the gradient difference threshold is set according to the overall gray distribution of the area, and a gray gradient distribution matrix is ​​generated.

[0035] Using the regional gray-level gradient values ​​in the gray-level gradient distribution matrix, the gray-level values ​​are adjusted through texture characteristics. Combining the weights of regional gray-level distribution and texture gradient characteristics, the gray-level difference values ​​are optimized using the following formula:

[0036]

[0037] Generate optimized grayscale values;

[0038] Among them, G t H represents the optimized grayscale value. x and H y W represents the grayscale difference between the horizontal and vertical directions. g and S gP represents the characteristic weight parameters in the horizontal and vertical directions, respectively. t Here, E represents the gradient distribution characteristic value, k is the texture enhancement adjustment coefficient, and E is the texture enhancement adjustment coefficient. g Weights are distributed for texture characteristics;

[0039] The optimized grayscale value is called to analyze the texture features. The texture layer characteristics are refined by optimizing the grayscale difference value after the texture gradient is optimized. The layer characteristics are superimposed and the layer gradient is adjusted according to the texture enhancement rules to generate a skin texture enhancement structure.

[0040] As a further aspect of the present invention, the step of obtaining the texture optimization feature map specifically includes:

[0041] Based on the skin texture enhancement structure, the region is divided into partitions according to the texture distribution characteristics. The region is segmented by detecting the directional changes of the texture, and the characteristics are extracted based on the density between textures. The texture characteristic data is then recombined to generate a region characteristic value matrix.

[0042] Based on the aforementioned regional characteristic value matrix, adjacent regional characteristics are compared item by item, local values ​​of regional characteristic differences are extracted, the distribution of adjacent regional characteristics is adjusted, and overall regional characteristic data is smoothed to generate a global matching characteristic value matrix.

[0043] Based on the global matching feature value matrix, the weights of regional texture features are redistributed, the adjacent unit values ​​of the feature value distribution pattern are adjusted, global texture features are optimized and reconstructed, and texture optimization feature maps are obtained.

[0044] As a further aspect of the present invention, the steps for obtaining the skin moisture distribution classification results are specifically as follows:

[0045] Extract the local region moisture content value and global distribution characteristics from the texture optimization feature map, and match the moisture content value with the local region coordinate parameters to generate the local region average moisture content and global moisture distribution characteristics.

[0046] The average local moisture content is classified and matched with the global moisture distribution characteristics. The classification is based on the differences between the average local moisture content and the global distribution characteristics, using the following formula:

[0047]

[0048] Calculate the weights of the differences between local and global moisture distribution to generate moisture distribution levels;

[0049] Among them, W d M represents the weight of the difference between local and global water distribution. r C represents the average moisture content of a local area. gQ represents the global water distribution characteristic parameter. t To match the weight coefficients, L u This is the variance parameter for the location of the local region;

[0050] The moisture distribution level is called, and the local area moisture distribution pattern is classified and compared according to the level. The consistency weight of moisture content and distribution characteristics between regions is matched to generate skin moisture distribution classification results.

[0051] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0052] In this invention, by precisely analyzing the boundaries, area, pixel values, and brightness values ​​of regions in a skin image, the moisture distribution status of each region can be effectively defined, and a weighted structure map can be generated for each region. This structured moisture information not only reflects the moisture differences between different regions but also provides accurate data support for further moisture analysis. During the moisture distribution modeling process, in-depth analysis of the relationships between nodes and their moisture value differences reveals the potential accumulation patterns of moisture in different parts of the skin, thus forming a highly accurate moisture distribution model. This model accurately displays the specific distribution characteristics of moisture on the skin surface, providing a reliable basis for subsequent texture optimization. The texture optimization process extracts gradient values ​​from areas of dense moisture and utilizes texture contrast enhancement technology to make the moisture differences in the image more significant, thereby presenting a more detailed skin moisture distribution map visually. Through final classification and grading, the processing flow not only accurately reflects the skin's moisture content but also clearly displays the moisture status of each region and performs differentiated analysis based on the moisture level of different regions. The overall solution, through layer-by-layer optimization, greatly improves the accuracy of moisture analysis and makes the formulation of skin care plans more scientific, with higher visualization effects and practical value. Attached Figure Description

[0053] Figure 1 This is a system flowchart of the present invention;

[0054] Figure 2 This is a flowchart of the region division results in this invention;

[0055] Figure 3 This is a flowchart of the skin region weighted structure in this invention;

[0056] Figure 4 This is a flowchart of the parameters governing the water distribution pattern in this invention;

[0057] Figure 5 This is a flowchart of the water distribution pattern model in this invention;

[0058] Figure 6 This is a flowchart of the skin texture enhancement structure in this invention;

[0059] Figure 7 This is a flowchart of the texture optimization feature map in this invention;

[0060] Figure 8 This is a flowchart illustrating the classification results of skin moisture distribution in this invention. Detailed Implementation

[0061] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0062] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0063] Please see Figure 1 The skin moisture analysis system based on computer vision technology includes:

[0064] The skin region segmentation module extracts the color and brightness values ​​of pixels based on the input skin image, analyzes the location and area of ​​region boundaries, summarizes the region adjacency relationship, establishes the region segmentation result, labels the weight of the water content range of region nodes, performs weighted analysis on the node connection relationship, and generates a weighted structure of skin regions.

[0065] The moisture distribution modeling module is based on the weighted structure of the skin region. It extracts the edge weight distribution values ​​between nodes, and performs moisture clustering operation according to the distribution pattern by superimposing the characteristic difference values ​​and the global characteristic values ​​to obtain the moisture distribution pattern parameters. It compares the differences in moisture values ​​between regions, analyzes the corresponding distribution relationship between regions through the difference values, and obtains the moisture distribution pattern model.

[0066] The texture enhancement and optimization module is based on the water distribution pattern model. It extracts the gray-level gradient values ​​of water-dense areas, adjusts the gray-level values ​​according to the texture gradient characteristics, optimizes the pixel difference values ​​in the area, performs texture contrast enhancement, refines the enhanced texture features in layers, generates a skin texture enhancement structure, and performs global matching on the regional characteristic values ​​to obtain the texture optimization feature map.

[0067] The moisture distribution classification module is based on texture optimization feature maps, extracts local area moisture content values ​​and global distribution characteristics, classifies the levels according to the range of regional moisture content, matches distribution pattern features, compares regions according to distribution levels, and generates skin moisture distribution classification results.

[0068] The region division results include boundary location, area, and adjacency relationship; the skin region weighting structure includes moisture content range and connectivity weight; the moisture distribution pattern parameters include distribution value differences and global characteristic analysis; the moisture distribution pattern model includes clustering operation results and characteristic difference analysis; the skin texture enhancement structure includes texture contrast enhancement and refinement, and feature layering; the texture optimization feature map includes grayscale value adjustment and global characteristic matching; and the skin moisture distribution classification results include moisture content level and matching distribution pattern features.

[0069] Please see Figure 2 The specific steps for obtaining the region division results are as follows:

[0070] Based on the input skin image, the color value and brightness value of each pixel in the skin image are extracted. A color distribution matrix is ​​constructed for the color value of the pixel, and a brightness distribution matrix is ​​constructed for the brightness value. The overall distribution range of color and brightness values ​​is analyzed based on the color distribution matrix and the brightness distribution matrix to obtain the preliminary distribution range parameters of color and brightness.

[0071] By scanning pixel by pixel, the RGB color and brightness values ​​of all pixels are recorded sequentially according to coordinate order. A color distribution matrix is ​​generated based on the pixel distribution of the extracted color values. The color distribution matrix contains the correspondence between pixel positions and color values. The main color distribution intervals are extracted by calculating the distribution density of the color histogram. A brightness distribution matrix is ​​generated for the brightness values, recording the brightness value and position of each pixel. The overall distribution characteristics of the brightness matrix are calculated by comparing the range of brightness distribution. Based on the distribution patterns of color values ​​and the distribution characteristics of brightness values, all pixels in the matrix are initially screened. Combining the statistical analysis of the distribution variation amplitude in the color histogram and the pixel proportion of each brightness interval in the brightness range, the pixel range that meets the distribution characteristics is screened through the joint calculation results of the color and brightness matrices. Finally, the preliminary color and brightness distribution range parameters are obtained.

[0072] Based on the preliminary color and brightness distribution range parameters, difference analysis is performed on the color and brightness values ​​of adjacent pixels. A threshold for color and brightness difference is set based on the calculation results. Boundary analysis is then performed on adjacent pixel pairs using the following formula:

[0073]

[0074] Calculate the boundary values ​​of pixels, mark pixels with values ​​greater than the threshold as boundary points, and generate a region boundary matrix;

[0075] Among them, B i,j C represents the boundary value of a pixel. i,j and C i+1,j L represents the color values ​​of adjacent pixels respectively. i,j and L i,j+1 W represents the brightness value of adjacent pixels respectively. c and W l These are the normalized weighting coefficients for color and brightness, respectively.

[0076] The advantage of the formula is that by calculating the color and brightness differences between adjacent pixels and combining them with normalized weight parameters, it comprehensively reflects the combined influence of color and brightness on the boundary, thereby enhancing the accuracy of boundary calculation and reducing the deviation caused by a single parameter.

[0077] Collect the color values ​​C of adjacent pixels in the color value matrix. i,j and C i+1,j And the brightness values ​​L of adjacent pixels in the brightness value matrix. i,j and L i,j+1 Calculate the difference;

[0078] Divide the color value difference and the brightness value difference by the weighting coefficient W respectively. c and W l Normalize the color and brightness values;

[0079] The normalized color value difference and brightness value difference are squared and summed, and the square root of the sum of squares is calculated to obtain the boundary value of the corresponding pixel in the boundary value matrix.

[0080] Let C i,j =100, C i+1,j =120, L i,j =80, L i,j+1 =60, W c =10, W l =20, substitute into the formula and calculate as follows:

[0081]

[0082] The results show that the boundary value of this pixel in the boundary value matrix is ​​2.236. If the boundary value is greater than the threshold, it is marked as a boundary point and further used to generate the boundary matrix.

[0083] Based on the region boundary matrix, the region is identified and segmented by performing connectivity analysis on the marked boundary points, the area of ​​the segmented region is counted, the shape parameters of each region are calculated based on the segmentation results, the adjacency relationship between regions is analyzed, and the region division results are established.

[0084] By comparing pixels one by one, pixels in the boundary matrix that are larger than a set threshold are marked. These marked pixels are recorded as a boundary marker set. Connectivity analysis is performed on the boundary marker set. Based on the connectivity between adjacent pixels, all pixels are divided into different regions. The area of ​​each region's pixel set is calculated by counting all pixels in the region. The area statistics are then used to extract shape parameters for each region. The shape parameter extraction calculates the perimeter of each region based on the area statistics and extracts the region's geometric features by combining the region's width and height ratio. The extracted geometric features are then compared with the region's adjacent pixel sets to establish the region division results and record the adjacency relationships of all regions.

[0085] Please see Figure 3 The specific steps for obtaining the weighted structure of the skin region are as follows:

[0086] Based on the regional division results, the range of regional nodes is extracted, and the moisture content data of nodes in the region is extracted. The upper and lower limits are analyzed, abnormal data is filtered, and the moisture data of edge nodes is corrected to obtain a table of regional node moisture content ranges.

[0087] The moisture content data of each node is analyzed sequentially. The initial data range of each node is grouped into fixed intervals, and the moisture distribution frequency of each interval is calculated one by one. The reliability of the node data is determined based on the frequency distribution characteristics. The upper and lower limits are analyzed. The moisture distribution density and mean of the regional nodes are calculated according to the upper and lower limits of the intervals. Multiple comparisons of the density and mean values ​​are used to determine whether there are any anomalies in the data. Anomalies are filtered out and removed from the calculation range. For the moisture content data of edge nodes, a weighted smoothing method is used to correct the edge data based on the data distribution pattern of neighboring nodes. The moisture data of neighboring nodes are proportionally affected by the weight allocation method. Finally, the correction value of the edge nodes is calculated. The moisture data of all nodes in the region are re-integrated to form a moisture content range table of regional nodes. The table clearly shows the upper and lower limits of the moisture content range of each node and the corresponding reliability assessment.

[0088] Based on the regional node moisture content range table, the upper and lower limits of node moisture content are normalized, the normalization weights are adjusted, the weight balance is optimized, the weight values ​​are recalibrated, and the regional node weighted table is obtained.

[0089] By mapping the upper and lower limits to a unified normalized interval, the normalized moisture value of each node is calculated, and the normalization weights are adjusted. Through data analysis on three aspects—the uniformity of node distribution across the entire region, the influence of neighboring nodes on the node, and the fluctuation characteristics of node moisture content—the normalized weight values ​​are redistributed to optimize weight balance. By verifying whether the node weights conform to the overall equilibrium principle within the region, and using the moisture distribution density of nodes within the region as a constraint, the weight distribution of individual nodes is adjusted to eliminate the interference of extreme weight values ​​on the overall analysis accuracy. The weight values ​​are then recalibrated. By comprehensively considering the spatial distribution characteristics and moisture content differences of regional nodes, the corrected weight values ​​are numerically fitted to ensure that the adjusted weights still effectively reflect the actual moisture characteristics of the nodes. Finally, a weighted table of regional nodes is formed, laying the data foundation for subsequent analysis.

[0090] Based on the weighted table of regional nodes, the connection relationship between nodes is analyzed, the geometric distance and normalized weight difference of moisture content between nodes are calculated, the weight matrix is ​​corrected, the node information and the weight matrix are integrated, and the weighted structure of skin region is generated.

[0091] By calculating the geometric distance between nodes, the spatial connection structure between nodes is determined one by one using triangular geometric relationships. The difference between the geometric distance and the normalized weight of moisture content between nodes is calculated, and the difference in moisture weight values ​​for each pair of nodes is analyzed differently. The difference in moisture content weight is used as an important parameter of the connection relationship. Combining the geometric distance and the weight difference, a preliminary correlation matrix between nodes is established. The weight matrix is ​​then corrected. Based on the balance index of the node weight distribution in the correlation matrix and the influence coefficient of the geometric distance, unreasonable weight differences in the matrix are corrected. The weights are redistributed through a weighted adjustment method to ensure that the matrix can accurately reflect the moisture distribution characteristics of the region. The node information and the weight matrix are integrated, and the moisture content data of the nodes are organically combined with the corrected weight matrix to further generate a weighted structure of the skin region for subsequent research and analysis of regional moisture distribution characteristics.

[0092] Please see Figure 4 The specific steps for obtaining parameters related to water distribution patterns are as follows:

[0093] Based on the skin region weighted structure, the edge weight values ​​between each node are extracted. By calculating the local characteristic difference values ​​between nodes and superimposing the global characteristic values, the edge weight values ​​are initially summarized. Combining the distribution characteristics of the edge weight values ​​between nodes with the summary results, the weighted node distribution matrix is ​​obtained.

[0094] By reading the initial edge weight values ​​between each node, the edge weight values ​​are standardized using the characteristic parameters of the skin region. The edge weight values ​​between all nodes are extracted and initially screened to remove outliers and invalid values. Then, all weight values ​​are normalized according to the standardization rules. The normalization process includes statistical analysis of the maximum and minimum values, calculating the normalization interval range through the difference, and correcting the normalized edge weights by combining the regional characteristic values ​​corresponding to the nodes. The correction process includes comparing the differences between the node characteristic values ​​and the edge weight distribution values, adjusting the edge weight values ​​according to the differences, and generating the final standardized edge weights. By analyzing the associated weight distribution of each node, the weighted node distribution matrix is ​​finally obtained.

[0095] Using the weight parameters recorded in the weighted node distribution matrix, a quantitative analysis of characteristic difference values ​​and global characteristic values ​​is performed, employing the following formula:

[0096]

[0097] Obtain the characteristic distribution difference matrix;

[0098] Among them, T k W represents the difference in characteristic distribution. a and W b D represents the local and global weight values, respectively. k For local characteristic values, G a is the global characteristic value, and n and m are the total number of local and global characteristics, respectively;

[0099] The advantage of the formula is that by weighting the differences between local and global characteristic values, the influence of local characteristic differences is amplified. At the same time, through the weight normalization process, the calculation results can accurately reflect the relationship between local and global characteristics, and improve the applicability of the results in subsequent analysis.

[0100] Collect the local characteristic values ​​D required for characteristic analysis k and the global property value set G a D is obtained by monitoring the local characteristic distribution values ​​of nodes in the skin region. k G was obtained through statistical analysis of the characteristic data of the entire skin area. a ;

[0101] Calculate the absolute value of the difference between local and global characteristic values, and use this as the core input for characteristic differences;

[0102] Introducing weight parameter W a and W b The weight values ​​are obtained by quantifying the importance of the features, where the local weight value W aThe squared value amplifies the effect of local differences by multiplying the difference value by the squared value of the corresponding weight.

[0103] All weighted difference values ​​are summed, and the sum is divided by the sum of the global weight values ​​∑W. b This enables normalization processing on a global scale.

[0104] Set local characteristic value D k =40, Global Characteristic Value G a =50, local weight value W a =0.7, global weight value W b =1.2, then the calculation is as follows:

[0105] |D k -G a |=|40-50|=10;

[0106]

[0107] ∑W b =1.2;

[0108]

[0109] The result shows that the characteristic distribution difference value is 4.083, indicating that there is a certain difference between the local characteristic value and the global characteristic value under the current calculation conditions. This value will be used as the core parameter for subsequent characteristic distribution difference analysis.

[0110] Based on the characteristic distribution difference matrix, clustering operations are performed on the nodes by analyzing the weighted weights between the nodes, and the water area characteristics are divided according to the node clustering pattern. The relationship between the node distribution pattern and the regional characteristics is extracted to obtain the water distribution pattern parameters.

[0111] The system retrieves the node weight distribution values ​​recorded in the matrix and performs classification and clustering analysis on the weight values ​​of each node. The analysis includes the range distribution of node weights, the differences in the associated weights between nodes, and the distribution of overall regional characteristic parameters. First, by extracting the weight parameters from the characteristic distribution difference matrix, all weight values ​​are sorted in ascending order. The distribution frequency of weight values ​​and the number of nodes within each interval are statistically analyzed. Then, the weight values ​​of each node are grouped according to the clustering rules. The classification rules are set based on the frequency changes of weight distribution and the differences in node characteristics within the weight range. By analyzing the node characteristics within each weight distribution range in the grouping results, the characteristic superposition parameters of each group are calculated. The characteristic superposition parameters include the total weight, the weighted average value between nodes, and the maximum weight value. The distribution pattern between groups is recorded. Through the analysis of the grouping characteristics between nodes, including the weight distribution characteristics of the water area and the grouping clustering relationship between nodes, the parameters of water distribution pattern are obtained.

[0112] Please see Figure 5 The specific steps for obtaining the water distribution pattern model are as follows:

[0113] Based on the parameters of water distribution pattern, the water content values ​​of nodes in the region are extracted, the node data are grouped by region, the water value distribution pattern is fitted, and a regional water distribution statistical table is generated.

[0114] The moisture content data of each node is obtained one by one. By combining the geographic location and moisture characteristics of each node, the distribution trend of node moisture data within the region is clarified. The node data is grouped by region, and the region is divided into multiple small sub-regions to ensure that the node data within each sub-region can reflect local moisture characteristics. The node data is integrated into independent groups by region through data aggregation and rearrangement. The distribution pattern of moisture value is fitted. Within each group, the mean, maximum, minimum and standard deviation of node moisture content are calculated. Based on the parameters, the moisture distribution curve of the region is fitted. For the distribution pattern of each region, the local moisture distribution trend is marked and a regional moisture distribution statistical table is generated. The distribution pattern parameters of all sub-regions are summarized into a table. The statistical table includes the node distribution range, mean moisture and its distribution characteristic parameters, so as to more comprehensively show the overall picture of moisture distribution in the skin area and provide reliable data support.

[0115] Based on the regional moisture distribution statistics table, the statistical parameters of regional moisture distribution are compared, the deviation of moisture distribution patterns between regions is analyzed, the correlation patterns between regions are integrated, and the data on the differences in regional moisture distribution are obtained.

[0116] By comparing the moisture distribution parameters in the statistical tables of various regions, the differences in mean moisture and standard deviation among regions are calculated. The range of differences between the maximum and minimum values ​​is analyzed step by step to extract the characteristic parameter differences of each region. The deviation of moisture distribution patterns among regions is analyzed. Combined with the spatial distribution characteristics of moisture data among regions, the deviation of moisture distribution trends in each region is calculated. According to the fitted curve of the distribution pattern, the moisture change trends among regions are compared to determine the main differences in moisture distribution patterns among regions. The correlation patterns among regions are integrated. By analyzing the overlap of node data between adjacent regions, the correlation of moisture characteristics between different regions is determined. The difference data and correlation information among regions are collected into a unified model to obtain the data on the differences in regional moisture distribution, providing an analytical basis for further exploration of the relationship between moisture distribution among regions.

[0117] Based on the data on regional water distribution differences, we extract the water distribution difference values ​​and correlation parameters between regions, collect the parameters into a unified model, map the regional difference values ​​to the parameter curve, integrate the water distribution parameters, and obtain a water distribution pattern model.

[0118] By comparing the differences in regional moisture data, this study analyzes the relevant parameters between regions one by one. Combining the moisture characteristics of nodes, key parameters related to moisture distribution are extracted, and the differences are grouped and organized. The parameters are then aggregated into a unified model, integrating the differential parameter data and related parameters between regions. The parameter data are standardized through a unified calculation model to ensure that the data from different regions have consistent comparative significance under the same model. Regional difference values ​​are mapped to parameter curves. For each group of parameter data, parameter curves are generated according to the changing trends of regional difference values. The curves reflect the correlation and difference characteristics of moisture distribution between regions. Moisture distribution parameters are integrated, and all parameters and difference values ​​are uniformly organized into the regional difference analysis model. Combining distribution patterns and difference relationships, a moisture distribution pattern model is generated, providing a visualized model foundation and accurate data support for the study of moisture distribution characteristics in skin regions.

[0119] Please see Figure 6 The specific steps for obtaining the skin texture enhancement structure are as follows:

[0120] Based on the water distribution pattern model, the gray values ​​of pixels in water-dense areas are extracted, the gray gradient values ​​are calculated, the gradient difference threshold is set according to the overall gray distribution of the area, and a gray gradient distribution matrix is ​​generated.

[0121] By analyzing the pixel grayscale matrix of densely water-distributed areas pixel by pixel, the grayscale value of each pixel is retrieved and its difference from the grayscale values ​​of its neighboring pixels is calculated. The grayscale difference value is obtained as the grayscale gradient value of the pixel using absolute value operation. Then, according to the overall grayscale characteristics of the densely water-distributed areas, the distribution range of grayscale gradient values ​​is analyzed, and a threshold range of grayscale difference is set. Grayscale gradient points exceeding the set threshold range are classified into high difference gradient ranges, and the remaining points are classified into low difference gradient ranges. By statistically analyzing the distribution ratio of high difference gradient points and low difference gradient points, a grayscale gradient distribution pattern map is generated. At the same time, combined with the overall change characteristics of the grayscale distribution in the region, the high difference range and the low difference range are summarized separately, and the grayscale value distribution of pixels in each range is recorded. Finally, a grayscale gradient distribution matrix is ​​generated. This matrix records the grayscale gradient value of each pixel and the grayscale characteristic distribution within the region for subsequent processing.

[0122] By utilizing the regional gray-level gradient values ​​in the gray-level gradient distribution matrix, and adjusting the gray-level values ​​through texture characteristics, the gray-level difference values ​​are optimized by combining the weights of regional gray-level distribution and texture gradient characteristics, using the following formula:

[0123]

[0124] Generate optimized grayscale values;

[0125] Among them, G t H represents the optimized grayscale value. x and Hy W represents the grayscale difference between the horizontal and vertical directions. g and S g P represents the characteristic weight parameters in the horizontal and vertical directions, respectively. t Here, E represents the gradient distribution characteristic value, k is the texture enhancement adjustment coefficient, and E is the texture enhancement adjustment coefficient. g Weights are distributed for texture characteristics;

[0126] The advantage of the formula is that by introducing gradient characteristic weights and texture enhancement adjustment coefficients, the gray-level gradients in the horizontal and vertical directions are uniformly normalized. At the same time, by superimposing characteristic weight values, the adaptability of the result to the texture distribution is improved, making the gray-level optimization result smoother and more adaptable to complex texture characteristics.

[0127] Obtain the horizontal gray-level difference value H from the gray-level gradient distribution matrix. x and vertical grayscale difference value H y The horizontal difference value is calculated by comparing the gray values ​​of adjacent horizontal pixels pixel by pixel, and the vertical difference value is calculated by comparing the gray values ​​of adjacent vertical pixels pixel by pixel.

[0128] The characteristic weight parameters W in the horizontal and vertical directions are determined based on the regional grayscale distribution characteristics. g and S g By analyzing the changes in the range of gray-level gradient distribution and texture distribution characteristic values ​​in the region, weight values ​​are calculated, and the texture enhancement adjustment coefficient k and texture characteristic distribution weight E are obtained. g ;

[0129] Substitute the above parameters into the formula, and normalize the grayscale difference values ​​in the horizontal and vertical directions. At the same time, obtain the grayscale optimization result by calculating the sum of squares.

[0130] Set the horizontal grayscale difference value H x =20, Vertical grayscale difference value H y =15, Horizontal characteristic weight W g =0.4, gradient distribution characteristic value P t =2.5, Vertical characteristic weight S g =0.6, texture enhancement adjustment coefficient k=2.0, texture feature distribution weight E g =10, calculated as follows:

[0131]

[0132] The results show that the optimized grayscale value is 45.2, which can be effectively used for further optimization of texture enhancement features.

[0133] The optimized grayscale values ​​are called to analyze the texture features. The texture layer characteristics are refined by optimizing the grayscale difference values ​​after the texture gradient is optimized. The layer characteristics are superimposed and the layer gradient is adjusted according to the texture enhancement rules to generate a skin texture enhancement structure.

[0134] The texture characteristic distribution of grayscale values ​​is analyzed and optimized pixel by pixel. The optimized grayscale gradient values ​​are extracted and divided into layers according to the gradient value range. The texture characteristic values ​​within the grayscale gradient range of each layer are called to generate corresponding texture characteristic layer matrices. Then, the characteristic analysis of each layer matrix is ​​combined to perform layer-by-layer cumulative analysis. The texture enhancement characteristics of the cumulative results are optimized. The optimization process includes the removal of abnormal gradient points in the texture characteristic matrix, the normalization adjustment of the texture characteristic range, and the smoothing of the cumulative gradient distribution. The overall texture characteristic distribution is optimized by superimposing the layer characteristic matrices. The optimized texture distribution results are called, and comprehensive analysis and induction operations are performed on the final texture characteristic matrix to ensure that the results can accurately record the enhancement characteristics of the regional texture distribution, and finally obtain the skin texture enhancement structure.

[0135] Please see Figure 7 The specific steps for obtaining the texture optimization feature map are as follows:

[0136] Based on the skin texture enhancement structure, the region is divided into partitions according to the texture distribution characteristics. The region is segmented by detecting the directional changes of the texture, and the characteristics are extracted based on the density between textures. The texture characteristic data is then recombined to generate a region characteristic value matrix.

[0137] By calculating the direction vector of texture distribution within each pixel region, the changes in pixel intensity gradient are analyzed point by point to clarify the texture direction trend of each small region. Segmentation is performed by detecting changes in the texture direction in the region. A fixed step size sliding analysis window is used to statistically analyze the principal vector of texture direction within each window and compare the direction differences between adjacent windows. Regions with significant direction changes are used as segmentation boundaries to complete the directional partitioning of the entire region. Characteristics are extracted based on the density between textures. By calculating the distribution density of texture intersection points within a unit region, the density characteristic value of texture in each region is extracted. Combined with the repetitive features of local textures, the principal frequency value and amplitude characteristic value of texture are further extracted. Texture characteristic data are reorganized. By representing the texture directionality, density, principal frequency value, and other characteristics of each partition in numerical form, the extracted characteristics are arranged in order of region to finally form a region characteristic value matrix, which serves as the core data input for subsequent skin texture analysis.

[0138] Based on the regional characteristic value matrix, the characteristics of adjacent regions are compared item by item, local values ​​of regional characteristic differences are extracted, the distribution of characteristics of adjacent regions is adjusted, the regional characteristic data is smoothed as a whole, and a global matching characteristic value matrix is ​​generated.

[0139] The texture direction, density, and dominant frequency of adjacent regions in the matrix are compared item by item. The absolute difference of each characteristic item is calculated, and the cumulative sum of the differences is used as a quantitative indicator of regional characteristic differences. The main characteristic difference points between each region are identified, and local values ​​of regional characteristic differences are extracted. The difference values ​​of regional characteristic differences that are higher than the global mean are extracted as key data for local adjustment. At the same time, the boundary of the region with the maximum characteristic difference is recorded for subsequent adjustment. The characteristic distribution of adjacent regions is adjusted by interpolating new characteristic values ​​at the boundary of adjacent regions to smooth the abrupt transition of the characteristic distribution curve, making the characteristic distribution between regions more continuous and reducing the analysis error caused by abrupt changes in texture characteristics. The regional characteristic data is smoothed as a whole. The characteristic values ​​of all regions are weighted and averaged according to density, direction, and frequency characteristics. The weight is assigned as the reciprocal of the texture similarity of adjacent regions. A global matching characteristic value matrix is ​​constructed with the smoothed characteristic values ​​to provide unified global data support for subsequent texture optimization operations.

[0140] Based on the global matching feature value matrix, the weights of regional texture features are redistributed, the adjacent unit values ​​of the feature value distribution pattern are adjusted, and the global texture features are optimized and reconstructed to obtain the texture optimization feature map.

[0141] By analyzing the proportion of texture characteristics in each region of the global matching feature value matrix, the influence of regional characteristics on the overall texture distribution is calculated. Weights are then redistributed according to this influence, adjusting the distribution pattern of adjacent unit values. Based on the value of each unit in the global matching feature value matrix, interpolation is performed on the feature values ​​of adjacent units to adjust their distribution pattern, ensuring a gradual transition of feature values ​​between adjacent units in the matrix. This avoids sudden changes in local values ​​and ensures overall consistency of texture characteristics. Global texture characteristic optimization and reconstruction are then performed. The adjusted feature values ​​are used as the basis for optimization, and the texture characteristic parameters of each region are recalculated using weights. The directional distribution, density distribution, and frequency characteristic distribution curves of the regional texture are reconstructed. The optimized features are stored in matrix form to obtain texture optimization feature maps. The final feature matrix is ​​then overlaid with the original skin texture image to demonstrate the spatial distribution and optimization results of skin surface texture characteristics, providing a precise reference for subsequent moisture analysis and skin quality assessment.

[0142] Please see Figure 8 The specific steps for obtaining the skin moisture distribution classification results are as follows:

[0143] Extract the local moisture content value and global distribution characteristics from the texture optimization feature map, and match the moisture content value with the local coordinate parameters to generate the local average moisture content and global moisture distribution characteristics.

[0144] By scanning the extracted texture optimization feature map pixel by pixel and performing grayscale conversion on each pixel value, the moisture value of each pixel is uniformly quantized to a grayscale value range of 0 to 255. The local area is divided into multiple grid units, each containing a certain number of pixels. Then, statistical analysis is performed on the pixels in each grid to calculate the mean moisture value of the local area. The mean is calculated by the arithmetic mean of the grayscale values ​​of all pixels. The local moisture distribution characteristics are completed by analyzing the local differences in the mean moisture value. Using the center point coordinates of each area as a reference, the mean deviation between adjacent areas is calculated, and the standard deviation of the mean moisture distribution in the global range is calculated. The standard deviation is obtained by calculating the dispersion of the mean moisture value in the local area. Then, the mean moisture value and standard deviation of the local area are combined to complete the normalization process of the global moisture distribution characteristics. The normalization results are organized into a table for further analysis, and finally, the mean moisture content of the local area and the global moisture distribution characteristics are generated.

[0145] The local average moisture content is classified and matched with the global moisture distribution characteristics. The classification is based on the differences between the local average moisture content and the global distribution characteristics, using the following formula:

[0146]

[0147] Calculate the weights of the differences between local and global moisture distribution to generate moisture distribution levels;

[0148] Among them, W d M represents the weight of the difference between local and global water distribution. r C represents the average moisture content of a local area. g Q represents the global water distribution characteristic parameter. t To match the weight coefficients, L u This is the variance parameter for the location of the local region;

[0149] The advantage of the formula is that by combining comprehensive parameters such as regional average moisture, global characteristics, and location differences, it refines the calculation of moisture distribution weights, which can reflect the degree of correlation between local areas and global characteristics, thereby improving the accuracy of classification results.

[0150] M r This represents the average moisture content of a local area, calculated based on the average pixel values ​​of that area. For example, if the pixel values ​​of a certain local area are [50, 60, 70, 80], then M is calculated. r = (50 + 60 + 70 + 80) / 4 = 65;

[0151] C g This represents the global moisture distribution characteristics, calculated using the statistical standard deviation of the mean values ​​across all regions. For example, if the region mean is [65, 70, 75], then C is calculated.g = (|65-70|+|70-75|+|75-65|) / 3=5;

[0152] Q t This represents the matching weight coefficient, calculated by normalizing the ratio of the number of pixels covered in the current local region to the maximum number of pixels covered. For example, if the number of pixels covered is 150, the maximum number of pixels covered is 200, and the minimum number of pixels covered is 100, then Q is calculated. t = (150-100) / (200-100) = 0.5;

[0153] L u This represents the variance of pixel positions within a local region. It is used to calculate the degree of dispersion within the region, combined with the pixel coordinates. For example, given coordinates [(1, 1), (1, 2), (2, 1), (2, 2)], calculate L. u =0.5;

[0154] Substitute the above parameters into the formula

[0155] The results indicate that the weight of the difference between local and global moisture distribution characteristics is 32.25. This weight is used for further moisture distribution level classification and matching, providing accurate characteristic parameters for subsequent distribution analysis.

[0156] The system calls up the moisture distribution level, classifies and compares the moisture distribution patterns of local areas according to the level, matches the consistency weight of moisture content and distribution characteristics between areas, and generates skin moisture distribution classification results.

[0157] Based on the classification results of local area moisture content values ​​and their corresponding distribution pattern characteristics, a parameter form for the matching pattern is constructed. By classifying the moisture content values ​​of local areas with their corresponding distribution levels, local areas within the same distribution level are compared, and the deviation between the moisture content values ​​of each area and the standard distribution level is calculated. The deviation is obtained by subtracting the mean of the standard level from the current local area moisture value and taking the absolute value. For the moisture distribution characteristics between matching areas, the consistency weight is calculated based on the distribution level of adjacent areas. The consistency weight is obtained by the proportion of overlapping pixels in the regional moisture distribution levels. Further combining the distribution statistics of the consistency weight between regions, and according to the comparison results of the distribution levels, an overall distribution model is constructed. Finally, the consistency between the local area moisture distribution pattern and the global characteristics is matched to generate the skin moisture distribution classification result.

[0158] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A skin moisture analysis system based on computer vision technology, characterized in that, The system includes: The skin region segmentation module extracts the color and brightness values ​​of pixels based on the input skin image, analyzes the location and area of ​​region boundaries, summarizes the region adjacency relationship, establishes the region segmentation result, labels the weight of the water content range of region nodes, performs weighted analysis on the node connection relationship, and generates a weighted structure of skin regions. The moisture distribution modeling module extracts the edge weight distribution values ​​between nodes based on the weighted structure of the skin region. By superimposing the characteristic difference values ​​and the global characteristic values, it performs moisture clustering operations according to the distribution pattern to obtain moisture distribution pattern parameters. It compares the differences in moisture values ​​between regions and analyzes the corresponding distribution relationship between regions through the difference values ​​to obtain a moisture distribution pattern model. The texture enhancement and optimization module extracts grayscale gradient values ​​in areas with dense moisture based on the moisture distribution pattern model, adjusts grayscale values ​​according to texture gradient characteristics, optimizes pixel differences within the area, performs texture contrast enhancement, refines the enhanced texture features in layers, generates a skin texture enhancement structure, and performs global matching of regional characteristic values ​​to obtain a texture optimization feature map. Based on the texture optimization feature map, the moisture distribution classification module extracts the local area moisture content value and global distribution characteristics, classifies the area moisture content range into levels, matches the distribution pattern features, compares the distribution levels by region, and generates skin moisture distribution classification results.

2. The skin moisture analysis system based on computer vision technology according to claim 1, characterized in that, The specific steps for obtaining the region division results are as follows: Based on the input skin image, the color value and brightness value of each pixel in the skin image are extracted. A color distribution matrix is ​​constructed for the color value of the pixel, and a brightness distribution matrix is ​​constructed for the brightness value. The overall distribution range of color and brightness values ​​is analyzed based on the color distribution matrix and the brightness distribution matrix to obtain the preliminary distribution range parameters of color and brightness. Based on the preliminary color and brightness distribution range parameters, difference analysis is performed on the color and brightness values ​​of adjacent pixels. A color and brightness difference threshold is set based on the calculation results. Boundary analysis is then performed on adjacent pixel pairs using the following formula: Calculate the boundary values ​​of pixels, mark pixels with values ​​greater than the threshold as boundary points, and generate a region boundary matrix; Among them, B i,j C represents the boundary value of a pixel. i,j and C i+1,j L represents the color values ​​of adjacent pixels respectively. i,j and L i,j+1 W represents the brightness value of adjacent pixels respectively. c and W l These are the normalized weighting coefficients for color and brightness, respectively. Based on the region boundary matrix, the regions are identified and segmented by performing connectivity analysis on the marked boundary points, the area of ​​the segmented regions is counted, the shape parameters of each region are calculated based on the segmentation results, the adjacency relationships between regions are analyzed, and the region division results are established.

3. The skin moisture analysis system based on computer vision technology according to claim 2, characterized in that, The specific steps for obtaining the weighted structure of the skin region are as follows: Based on the region division results, the range of regional nodes is extracted, and the moisture content data of the nodes in the region is extracted. The upper and lower limits are analyzed, abnormal data is filtered, and the moisture data of edge nodes is corrected to obtain a table of the range of regional node moisture content. Based on the regional node moisture content range table, the upper and lower limits of node moisture content are normalized, the normalization weights are adjusted, the weight balance is optimized, the weight values ​​are recalibrated, and the regional node weighted table is obtained. Based on the weighted table of regional nodes, the node connection relationship is analyzed, the geometric distance and normalized weight difference of moisture content between nodes are calculated, the weight matrix is ​​corrected, the node information and the weight matrix are integrated, and a weighted structure of the skin region is generated.

4. The skin moisture analysis system based on computer vision technology according to claim 3, characterized in that, The specific steps for obtaining the moisture distribution pattern parameters are as follows: Based on the weighted structure of the skin region, the edge weight values ​​between each node are extracted. By calculating the local characteristic difference values ​​between nodes and superimposing the global characteristic values, the edge weight values ​​are initially summarized. Combining the distribution characteristics of the edge weight values ​​between nodes with the summary results, the weighted node distribution matrix is ​​obtained. Using the weight parameters recorded in the weighted node distribution matrix, a quantitative analysis of the characteristic difference value and the global characteristic value is performed, using the formula: Obtain the characteristic distribution difference matrix; Among them, T k W represents the difference in characteristic distribution. a and W b D represents the local and global weight values, respectively. k For local characteristic values, G a is the global characteristic value, and n and m are the total number of local and global characteristics, respectively; Based on the aforementioned characteristic distribution difference matrix, clustering operations are performed on the nodes by analyzing the weighted weights between the nodes, and the water area characteristics are divided according to the node clustering pattern. The relationship between the node distribution pattern and the regional characteristics is extracted to obtain the water distribution pattern parameters.

5. The skin moisture analysis system based on computer vision technology according to claim 4, characterized in that, The specific steps for obtaining the moisture distribution pattern model are as follows: Based on the aforementioned moisture distribution pattern parameters, the moisture content values ​​of nodes within the region are extracted, the node data are grouped by region, the moisture value distribution pattern is fitted, and a regional moisture distribution statistical table is generated. Based on the aforementioned regional moisture distribution statistics table, the statistical parameters of regional moisture distribution are compared, the deviations in moisture distribution patterns between regions are analyzed, the correlation patterns between regions are integrated, and the data on regional moisture distribution differences are obtained. Based on the data on regional water distribution differences, the differences in water distribution between regions and the correlation parameters are extracted, the parameters are aggregated into a unified model, the regional differences are mapped to parameter curves, the water distribution parameters are integrated, and a water distribution pattern model is obtained.

6. The skin moisture analysis system based on computer vision technology according to claim 5, characterized in that, The specific steps for obtaining the skin texture enhancement structure are as follows: Based on the moisture distribution pattern model, the pixel gray values ​​in the densely moist areas are extracted, the gray gradient values ​​are calculated, the gradient difference threshold is set according to the overall gray distribution of the area, and a gray gradient distribution matrix is ​​generated. Using the regional gray-level gradient values ​​in the gray-level gradient distribution matrix, the gray-level values ​​are adjusted through texture characteristics. Combining the weights of regional gray-level distribution and texture gradient characteristics, the gray-level difference values ​​are optimized using the following formula: Generate optimized grayscale values; Among them, G t H represents the optimized grayscale value. x and H y W represents the grayscale difference between the horizontal and vertical directions. g and S g P represents the characteristic weight parameters in the horizontal and vertical directions, respectively. t Here, E represents the gradient distribution characteristic value, k is the texture enhancement adjustment coefficient, and E is the texture enhancement adjustment coefficient. g Weights are distributed for texture characteristics; The optimized grayscale value is called to analyze the texture features. The texture layer characteristics are refined by optimizing the grayscale difference value after the texture gradient is optimized. The layer characteristics are superimposed and the layer gradient is adjusted according to the texture enhancement rules to generate a skin texture enhancement structure.

7. The skin moisture analysis system based on computer vision technology according to claim 6, characterized in that, The specific steps for obtaining the texture optimization feature map are as follows: Based on the skin texture enhancement structure, the region is divided into partitions according to the texture distribution characteristics. The region is segmented by detecting the directional changes of the texture, and the characteristics are extracted based on the density between textures. The texture characteristic data is then recombined to generate a region characteristic value matrix. Based on the aforementioned regional characteristic value matrix, adjacent regional characteristics are compared item by item, local values ​​of regional characteristic differences are extracted, the distribution of adjacent regional characteristics is adjusted, and overall regional characteristic data is smoothed to generate a global matching characteristic value matrix. Based on the global matching feature value matrix, the weights of regional texture features are redistributed, the adjacent unit values ​​of the feature value distribution pattern are adjusted, global texture features are optimized and reconstructed, and texture optimization feature maps are obtained.

8. The skin moisture analysis system based on computer vision technology according to claim 7, characterized in that, The specific steps for obtaining the skin moisture distribution classification results are as follows: Extract the local region moisture content value and global distribution characteristics from the texture optimization feature map, and match the moisture content value with the local region coordinate parameters to generate the local region average moisture content and global moisture distribution characteristics. The average local moisture content is classified and matched with the global moisture distribution characteristics. The classification is based on the differences between the average local moisture content and the global distribution characteristics, using the following formula: Calculate the weights of the differences between local and global moisture distribution to generate moisture distribution levels; Among them, W d M represents the weight of the difference between local and global water distribution. r C represents the average moisture content of a local area. g Q represents the global water distribution characteristic parameter. t To match the weight coefficients, L u This is the variance parameter for the location of the local region; The moisture distribution level is called, and the local area moisture distribution pattern is classified and compared according to the level. The consistency weight of moisture content and distribution characteristics between regions is matched to generate skin moisture distribution classification results.

Images