Skin visual detection method based on LED lamp

Abstract

The invention relates to the technical field of skin detection, and particularly discloses a skin visual detection method based on an LED lamp, and the method comprises the steps: obtaining a multiband LED light source, and generating multiband detection light based on the multiband LED light source; acquiring reflection spectrum data and filtered scattered light data of the skin based on the multiband detection light, and generating image information to be recognized according to the reflection spectrum data; acquiring skin pigment information of the skin surface according to the to-be-recognized image information, and generating real-time skin modal feature data according to the skin pigment information; acquiring skin specular reflection characteristics and skin diffuse reflection characteristics of the skin surface according to the filtered scattered light data, and generating skin reflected light data according to the skin specular reflection characteristics and the skin diffuse reflection characteristics. The multi-band LED light source is used for generating multi-band detection light, and the skin state is comprehensively evaluated by obtaining reflection spectrum data and filtered scattered light data of the skin.

Description

Technical Field

[0001] This invention relates to the field of skin detection technology, and in particular to a skin visual detection method based on LED lights. Background Technology

[0002] With the continuous advancement of modern technology, skin testing technology has been widely applied in fields such as medicine, beauty, and health management. Skin condition assessment evaluates the health status of the skin by detecting its visual characteristics, and its application in detecting skin pigmentation, texture, and elasticity is particularly important.

[0003] Traditional methods rely on a single light source, such as blue or red light, which can only acquire single-dimensional information about skin color or surface texture. For example, blue light alone cannot distinguish the spectral differences between melanin and hemoglobin, leading to an increased misdiagnosis rate of melasma and post-inflammatory hyperpigmentation. This is because the spectral coverage of a single light source is limited and cannot fully stimulate the optical response of skin tissue. When using blue light alone, melanin's preset broadband absorption characteristics overlap with hemoglobin's preset characteristic absorption peaks, making the reflectance spectral data unable to accurately reflect the true composition of skin tissue. Furthermore, a single light source cannot capture the optical characteristics of deep skin structures, such as the scattering effect of dermal spots and the reflectance difference of epidermal pigments, leading to deviations in the judgment of lesion depth. Summary of the Invention

[0004] The purpose of this invention is to provide a skin visual detection method based on LED lights to solve the technical problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A skin visual detection method based on LED lights includes: Acquire a multi-band LED light source and generate multi-band detection light based on the multi-band LED light source; Based on the multi-band detection light, the skin's reflectance spectrum data and filtered scattered light data are acquired, and the image information to be identified is generated based on the reflectance spectrum data; Based on the image information to be identified, obtain skin pigment information on the skin surface, and generate real-time skin modality feature data based on the skin pigment information; Based on the filtered scattered light data, obtain the skin specular reflection characteristics and skin diffuse reflection characteristics of the skin surface, and generate skin reflected light data based on the skin specular reflection characteristics and skin diffuse reflection characteristics; Skin condition information is obtained based on the real-time skin modal feature data and skin reflectance data, and skin defect information is obtained based on the skin condition information; A skin condition assessment report is generated based on the skin defect information.

[0006] Preferably, the step of generating multi-band detection light based on the multi-band LED light source includes: The multi-band LED light source generates a multi-spectral illumination band, wherein the multi-spectral illumination band sequentially activates the ultraviolet, blue, green, and red light bands; An angular diffusion process is performed on the multispectral illumination band, and the multispectral diffused illumination band after diffusion processing is obtained. Obtain the ambient light intensity in real time; Based on the established mapping relationship between multispectral diffused illumination bands and ambient light intensity, the illumination-environment mapping correlation degree is obtained; The multi-band LED light source is corrected based on the lighting-environment mapping correlation to obtain the multi-band detection light.

[0007] Preferably, the step of generating image information to be identified based on the reflectance spectral data includes: Based on the reflected spectral data, multi-band reflected spectral data of skin reflection is obtained, wherein the multi-band reflected spectral data includes UV light signals and visible light reflected signals; Based on the multi-band reflectance spectral data, UV simulation images and reflected light simulation images are generated; Obtain the reflection position coordinate information of the simulated reflected light image; Obtain the real-time skin image corresponding to the reflection position coordinates; The generated UV simulation image, reflected light simulation image, and real-time skin image are aligned, and skin feature information is extracted; The image information to be identified is generated based on the skin feature information.

[0008] Preferably, the step of obtaining skin pigmentation information of the skin surface based on the image information to be identified includes: Pixel feature point information is extracted based on the image information to be identified; The skin reflectance ratio is obtained from the simulated image of reflected light. A melanin concentration-reflectance mapping relationship is established based on the pixel feature point information and the skin reflectance ratio. The melanin concentration content is obtained based on the melanin concentration-reflectance mapping relationship. The real-time contrast between the melanin deposition area and normal skin is obtained based on the pixel feature point information. A pigment index distribution map is generated based on the real-time contrast. Skin pigmentation information on the skin surface is obtained based on the pigment index distribution map.

[0009] Preferably, the step of generating real-time skin modal feature data based on skin pigmentation information includes: The location information of raised areas and recessed areas on the skin surface is obtained based on the real-time skin image. The skin surface roughness is obtained based on the location information of the protruding and concave areas. Based on the skin pigment information, obtain pigment grading texture information, and extract skin texture features based on the pigment grading texture information; A roughness-texture correlation mapping is established based on the skin surface roughness and skin texture features, and the roughness-texture correlation degree is obtained; Skin elasticity characteristics are obtained based on the roughness-texture correlation. Real-time skin modal feature data is generated based on the skin elasticity characteristics.

[0010] Preferably, the step of obtaining the skin specular reflection characteristics and skin diffuse reflection characteristics of the skin surface based on the filtered scattered light data includes: A reflected light image in a preset direction is obtained based on the filtered scattered light data; The specular reflection angle set and diffuse reflection angle set of the skin surface are obtained based on the reflected light image; The spatial distribution of the specular reflection region and the spatial distribution of the diffuse reflection region are obtained based on the specular reflection angle set and the diffuse reflection angle set. The skin's specular reflection characteristics are obtained based on the spatial distribution of the specular reflection areas. The diffuse reflection characteristics of the skin are obtained based on the spatial distribution of the diffuse reflection region.

[0011] Preferably, the step of generating skin reflectance data based on the skin specular reflection characteristics and the skin diffuse reflection characteristics includes: The specular and diffuse reflection components are obtained based on the skin specular and diffuse reflection characteristics, and the reflectance ratio is obtained based on the specular and diffuse reflection components. The reflectance of the micro-convex structure and the reflectance of the texture boundary on the skin surface are obtained based on the reflectance ratio. Skin texture sharpness is obtained based on the reflectivity of the micro-convex structure and the reflectivity of the texture boundary. A skin feature synthesis image is obtained based on the skin texture sharpness and the real-time skin image; Skin reflectance data is generated from the synthesized image based on the skin features.

[0012] Preferably, the step of obtaining skin state information based on the real-time skin modal feature data and skin reflectance data, and obtaining skin defect information based on the skin state information, includes: Based on the real-time skin modality feature data, pigment concentration, texture roughness, and skin condition depth are obtained; The texture reflectance is obtained based on the skin reflectance data; Skin condition information is obtained based on the skin pigment concentration, texture roughness, skin condition depth, and texture reflectivity. Multiple skin segmentation regions are obtained based on the skin condition information; A skin condition assessment is obtained for each skin segmentation region, and it is determined whether the skin condition assessment is less than a preset threshold. If it is less than 1 / 3, the skin segmentation area is judged to be in a healthy state. If it is greater than or equal to, the skin segmentation area is judged to be in an unhealthy state; Information on skin defects is obtained by segmenting areas of unhealthy skin.

[0013] Preferably, the step of generating a skin condition assessment report based on the skin defect information includes: Based on the skin defect information, obtain the defect pigment ratio, defect area ratio, and defect severity; The defect index is obtained based on the defect pigment ratio, defect area ratio, and severity. Based on the defect index, a skin condition assessment report is generated according to the defect pigment ratio, defect area ratio, and defect severity.

[0014] The beneficial effects of this application are as follows: This invention utilizes a multi-band LED light source to generate multi-band detection light, and achieves a comprehensive assessment of skin condition by acquiring skin reflectance spectrum data and filtered scattered light data. This method can accurately acquire pigment information, texture features, and reflectance data of the skin surface, thereby generating real-time skin modal feature data and reflectance data, and further acquiring skin condition information and skin defect information, ultimately forming a skin condition assessment report. It enables real-time and accurate detection of the skin, effectively assessing not only the health status of the skin but also identifying skin defects and problems. This method fully utilizes the advantages of multi-band light sources, ensuring the accuracy of the detection light and improving the reliability of the detection results through angle diffusion processing and correction of ambient light intensity. Simultaneously, this invention combines multiple data such as reflectance spectrum data, skin pigment concentration, and texture roughness, making the detection method more comprehensive and able to reflect the multi-dimensional characteristics of the skin. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of a method flow according to an embodiment of this application. Detailed Implementation

[0016] like Figure 1As shown, this application provides a skin visual detection method based on LED lights, including: S1. Acquire a multi-band LED light source and generate multi-band detection light based on the multi-band LED light source; S2. Based on the multi-band detection light, acquire the skin's reflectance spectrum data and filtered scattered light data, and generate the image information to be identified based on the reflectance spectrum data; S3. Obtain skin pigment information on the skin surface based on the image information to be identified, and generate real-time skin modality feature data based on the skin pigment information; S4. Obtain the skin specular reflection characteristics and skin diffuse reflection characteristics of the skin surface based on the filtered scattered light data, and generate skin reflected light data based on the skin specular reflection characteristics and skin diffuse reflection characteristics; S5. Obtain skin state information based on the real-time skin modal feature data and skin reflective light data, and obtain skin defect information based on the skin state information; S6. Generate a skin condition assessment report based on the skin defect information.

[0017] As described in steps S1-S6 above, traditional skin detection methods often rely on a single light source or do not consider the influence of ambient light intensity, resulting in inaccurate detection results or significant interference from environmental factors. However, the multi-band LED light source combined with dynamic correction of ambient light intensity proposed in this invention can effectively avoid these problems and improve the stability of the light source and the accuracy of detection. This invention generates detection light of different wavelengths using multi-band LED light sources, covering a variety of spectra including ultraviolet, blue, green, and red light. These different wavelengths of light can stimulate different optical responses on the skin surface, capturing the reflective properties of different layers of the skin surface and interior. Different wavelengths of light can provide multi-dimensional skin information, such as pigment distribution, texture, and health status. Furthermore, by covering different wavelengths, the multi-band LED light source can more comprehensively analyze various skin characteristics, avoiding information loss or false detections that may occur with a single light source.

[0018] Subsequently, the spectral data of skin reflection under multi-band light illumination can reveal the skin's pigment concentration, texture, and roughness characteristics. By filtering the scattered light, we can effectively remove environmental interference factors, making the acquired skin reflection data more accurate. Based on this reflection data, image information to be identified is generated. By combining the reflection spectral data and filtered scattered light data, the influence of ambient light on the measurement results can be eliminated, improving the accuracy of image recognition and obtaining more refined skin condition information, avoiding the deviations caused by changes in ambient light in traditional methods. Then, by analyzing the pigment information in the reflectance image, the distribution and concentration of melanin in the skin can be accurately identified, thereby determining the skin's health status. Real-time skin modal feature data is generated by combining multiple dimensions of pigment information, texture information, and skin elasticity. By generating skin modal feature data in real time, it can dynamically reflect changes in skin condition, especially under different lighting conditions, and can better capture subtle changes in the skin, greatly improving the accuracy of skin health monitoring. Next, specular reflection and diffuse reflection are two important optical properties of the skin surface. Specular reflection reflects the smoothness of the skin surface, while diffuse reflection reveals the roughness and texture of the skin. By extracting these features, a comprehensive understanding of the physical structure of the skin surface can be obtained. Through detailed analysis of specular and diffuse reflection characteristics, this invention can accurately distinguish between healthy skin conditions and potential lesion areas in skin detection. Subsequently, by integrating skin modal feature data and reflected light data, the system can comprehensively assess the skin's health status, including texture, pigmentation, skin moisture, and elasticity. Based on this data, the system can effectively identify blemishes or potential lesions on the skin, promptly detect skin problems, and comprehensively analyze skin modal feature data and reflected light data to more accurately and comprehensively assess the skin's health status. Finally, based on the skin defect information obtained in step S5, a detailed skin condition assessment report is generated. This report not only includes the skin's health status but also provides information on the defect pigmentation ratio, defect area percentage, and defect severity. Through the technical implementation of the above steps, this invention provides an innovative and comprehensive skin detection method that can efficiently and accurately acquire multi-dimensional features of the skin, solving the problems of unstable light source, environmental interference, and inaccurate skin detection in traditional technologies.

[0019] In one embodiment, the step of generating multi-band detection light based on the multi-band LED light source includes: S101. Generate a multispectral illumination band based on the multi-band LED light source, wherein the multispectral illumination band sequentially activates the ultraviolet, blue, green, and red light bands; S102. Perform angular diffusion processing on the multispectral illumination band and obtain the multispectral diffused illumination band after diffusion processing. S103. Obtain the ambient light intensity of the real-time environment; S104. Based on the established mapping relationship between the multispectral diffused illumination band and the ambient light intensity, the illumination-environment mapping correlation degree is obtained; S105. Based on the lighting-environment mapping correlation, the multi-band LED light source is corrected to obtain the multi-band detection light.

[0020] As described in steps S101-S105 above, this invention generates multi-band detection light using a multi-band LED light source and optimizes the effect of the light source through a correction process. By mapping and correcting the relationship between multi-spectral illumination and ambient light, the illumination accuracy of the light source is improved.

[0021] Firstly, while LED light sources are commonly used in lighting applications, their single-band nature often fails to meet the demands of multispectral detection. By activating multiple different bands (ultraviolet, blue, green, and red), more diverse lighting effects can be provided, adapting to different types of optical detection. This is achieved by activating different spectral bands in a multi-band LED light source, creating multiple spectral illumination bands. This step ensures the light source is suitable for multispectral applications, improving detection accuracy. Specifically, the ultraviolet, blue, green, and red light bands are activated step-by-step to create a broad-coverage multispectral illumination effect. The intensity and direction of illumination from different wavelengths of light sources will have different effects on the detection results. Angular diffusion processing can make the illumination of the light source uniformly distributed, thereby avoiding situations where the local light intensity is too high or too low, ensuring more accurate optical measurement and detection. Through angular diffusion processing, the multispectral bands of the light source are uniformly diffused, further optimizing the illumination effect and ensuring the uniformity of light throughout the detection area; Real-time acquisition of ambient light intensity ensures a response to real-time illumination changes, providing necessary data support for subsequent correction processes and preventing interference from ambient light on detection results. Then, by quantifying the mapping relationship between multispectral illumination bands and ambient light intensity, the illumination-environment mapping correlation can be accurately obtained, providing a more precise basis for light source correction and improving the system's stability and adaptability. The illumination-environment mapping relationship clearly defines the interaction between multispectral illumination bands and ambient light. By establishing this mapping relationship, the combined effect of illumination light and ambient light on the detection results can be quantified. Finally, by correcting the correlation between lighting and environment mapping, the output of the LED light source is optimized, enabling the light source to stably output multi-band detection light under different environments, achieving the goal of accurate detection. Through the correction of the mapping relationship, the multi-band output of the light source can be dynamically adjusted according to the ambient light intensity. This correction mechanism allows the system to still stably provide high-quality detection light even when the ambient light intensity changes.

[0022] In one embodiment, the step of generating image information to be identified based on the reflectance spectral data includes: S201. Obtain multi-band reflectance spectral data of skin reflection based on the reflectance spectral data, wherein the multi-band reflectance spectral data includes UV light signals and visible light reflectance signals; S202. Generate a UV simulation image and a reflected light simulation image based on the multi-band reflectance spectrum data; S203. Obtain the reflection position coordinate information of the simulated reflected light image; S204. Obtain the real-time skin image corresponding to the reflection position coordinate information; S205. Align the generated UV simulation image, reflected light simulation image, and real-time skin image, and extract skin feature information; S206. Generate image information to be identified based on the skin feature information.

[0023] As described in steps S201-S206 above, this invention acquires multi-band reflectance spectral information of the skin through reflectance spectral data, including ultraviolet (UV) light signals and visible light reflectance signals. Different bands of spectral reflectance signals can provide optical information about different layers of the skin. UV light signals primarily affect the surface reflectance characteristics of the skin, while visible light signals provide detailed information about skin color and texture. By combining UV and visible light signals, multi-band reflectance spectral data of the skin is acquired, thus comprehensively acquiring the reflectance characteristics of the skin surface and its deeper layers. UV simulation images can simulate the reflectance characteristics of the skin under ultraviolet light, while reflected light simulation images can represent the appearance of the skin under visible light. By generating UV simulation images and reflected light simulation images separately, the influence of ultraviolet and visible light can be clearly separated, thereby obtaining a more accurate skin feature image. This separation process allows the characteristics of each band to be represented independently. Using reflectance spectral data, a specialized image processing algorithm processes the reflectance signals of each band, converting them into corresponding simulation images. These images reflect the effects of different wavelengths of light on skin surface reflection. Through data mapping and image synthesis techniques, the final UV simulation image and reflected light simulation image are obtained. By extracting the reflection position coordinates, the spatial distribution of the skin surface reflection signal can be determined, providing precise positioning for image alignment and feature extraction. Subsequently, by acquiring real-time skin images corresponding to the reflection position coordinates, the consistency between the images and reflection data can be ensured, reducing errors caused by time differences. Through feature point matching and optical flow techniques, different images are spatially aligned, and skin texture and color feature information is extracted. This process ensures that corresponding feature points in each image are accurately matched, avoiding errors caused by image misalignment. By extracting skin feature information, different types of skin states or features can be effectively distinguished.

[0024] In one embodiment, the step of obtaining skin pigmentation information of the skin surface based on the image information to be identified includes: S301. Extract pixel feature point information based on the image information to be identified; S302. Obtain the skin reflectance ratio based on the simulated image of reflected light; S303. Establish a melanin concentration-reflection mapping relationship based on the pixel feature point information and the skin reflectance ratio; S304. Obtain the melanin concentration content based on the melanin concentration-reflection mapping relationship; S305. Obtain the real-time contrast between the melanin deposition area and normal skin based on the pixel feature point information; S306. Generate a pigment index distribution map based on the real-time contrast. S307. Obtain skin pigmentation information on the skin surface based on the pigment index distribution map.

[0025] As described in steps S301-S307 above, this invention extracts pixel feature point information from the image to be identified. Pixel feature points refer to points with specific meaning in an image, which play a crucial role in image analysis. Each pixel in an image carries physical information related to its position, color, and brightness. By extracting these feature points, the image can be decomposed as a whole. The skin reflectivity ratio is obtained from the simulated reflected light image. Skin reflectivity is a physical quantity describing the skin surface's ability to reflect light, and the reflectivity ratio reflects the light reflection characteristics of different areas of the skin surface. Under different lighting conditions, skin reflectivity exhibits certain regularities, especially the melanin content, which directly affects the skin's reflectivity characteristics. By simulating the interaction between light and skin, a simulated reflected light image is generated. By comparing the brightness differences between the original image and the simulated image, the skin reflectivity ratio is obtained. Specifically, using a reflective optics model, the reflectivity of each region is calculated based on the skin's material, texture, and melanin distribution, and their ratios are obtained. A mapping relationship between melanin concentration and reflectivity is established between pixel feature point information and the skin reflectivity ratio. Melanin concentration in the skin directly affects its light reflection characteristics. Through data acquisition and statistical analysis, the reflectivity ratios of different skin regions are extracted and their corresponding melanin concentrations are determined. Based on this, a regression analysis method is used to establish a melanin concentration-reflectivity mapping relationship. According to this relationship, the melanin concentration of the skin is obtained. Using the previously established mapping relationship between reflectivity and melanin concentration, the melanin concentration of each skin region can be directly calculated from the simulated reflected light image. Then, based on pixel feature point information, the real-time contrast between melanin-deposited areas and normal skin is obtained. The color difference between melanin-deposited areas and normal skin is significant; therefore, the contrast reflects the degree of melanin distribution. The contrast between melanin-deposited areas and normal skin is derived through the brightness differences of skin regions. Color difference information is extracted, and the contrast is updated in real time based on pixel feature point data. Using data visualization techniques, a pigment index distribution map is generated based on the acquired real-time contrast data. This map is then mapped to different regions of the image, using color differences to represent the melanin concentration in different areas. The pigment index distribution map visually displays the melanin distribution in various regions of the skin. Subsequently, pigment information on the skin surface is obtained based on the pigment index distribution map. Analysis of the pigment index distribution map provides an overall picture of skin pigment distribution.

[0026] In one embodiment, the step of generating real-time skin modality feature data based on skin pigmentation information includes: S308. Obtain the location information of the raised areas and the location information of the sunken areas on the skin surface based on the real-time skin image; S309. Obtain the skin surface roughness based on the location information of the protruding area and the location information of the concave area; S310. Obtain pigment gradation texture information based on the skin pigment information, and extract skin texture features based on the pigment gradation texture information; S311. Establish a roughness-texture correlation mapping based on the skin surface roughness and skin texture features, and obtain the roughness-texture correlation degree; S312. Obtain skin elasticity features based on the roughness-texture correlation. S313. Generate real-time skin modal feature data based on the skin elasticity characteristics.

[0027] As described in steps S308-S313 above, this invention utilizes real-time skin images, combined with advanced machine vision and image processing technologies, to accurately locate raised and recessed areas on the skin surface through feature extraction algorithms. This process uses high-resolution images and deep learning algorithms to segment and process the images, automatically identifying and labeling subtle changes on the skin surface. A deep convolutional neural network (CNN) is used to process the real-time skin images, and the trained model identifies local structural features of the skin to locate raised and recessed areas. Skin roughness is one of the important indicators for measuring skin condition. The more pronounced the raised and recessed areas, the higher the skin roughness. The magnitude of roughness directly affects the skin's luster and feel, and is also closely related to aging and environmental factors. By acquiring the relative height differences of raised and recessed areas, combined with the overall undulation pattern of the skin surface, the skin roughness is accurately obtained. Using the location information of raised and recessed areas, combined with texture analysis technology, and employing mathematical models (such as surface irregularity measurement methods), the degree of undulation of the skin surface is quantitatively obtained. Specifically, by acquiring the distance and height of local feature points in an image, the surface roughness of the skin is determined. Pigment distribution is a crucial indicator of skin health and aesthetics. Pigment information is closely related to skin health, the aging process, and the influence of the external environment. Analyzing the texture features of pigment distribution helps assess skin aesthetics, uniformity, and anti-aging capabilities. By combining skin pigment information and texture analysis algorithms, the gradation texture information of pigments can be accurately extracted. Using high-resolution images and deep learning models, the distribution of pigments on the skin is automatically identified, thereby generating texture feature data. Skin roughness and texture features are two interrelated factors that jointly reflect the skin's condition. By establishing a roughness-texture correlation mapping model and combining it with multidimensional data analysis techniques, more accurate skin feature predictions can be obtained. In this way, not only can the surface roughness of the skin be comprehensively considered, but also texture changes can be taken into account. Skin elasticity is one of the important indicators for assessing skin health and is closely related to aging, damage, and environmental pollution. Skin elasticity is closely related to roughness and texture. By combining the correlation between roughness and texture, advanced machine learning algorithms can accurately predict the elasticity characteristics of the skin. This method provides real-time feedback on skin elasticity, aiding in further assessment of skin health. Ultimately, it generates real-time skin modal feature data based on skin elasticity characteristics for a comprehensive evaluation of skin health. Through analysis of real-time skin images and combined with a multidimensional data model, it generates precise skin modal feature data. This data not only encompasses information from multiple dimensions of the skin surface but also reflects real-time changes in skin health.

[0028] In one embodiment, the step of obtaining the skin specular reflection features and skin diffuse reflection features of the skin surface based on the filtered scattered light data includes: S401. Obtain a reflected light image in a preset direction based on the filtered scattered light data; S402. Obtain the set of specular reflection angles and the set of diffuse reflection angles on the skin surface based on the reflected light image; S403. Obtain the spatial distribution of the specular reflection region and the spatial distribution of the diffuse reflection region based on the specular reflection angle set and the diffuse reflection angle set; S404. Obtain skin specular reflection characteristics based on the spatial distribution of the specular reflection area; S405. Obtain the diffuse reflection characteristics of the skin based on the spatial distribution of the diffuse reflection region.

[0029] As described in steps S401-S405 above, this invention obtains the skin's specular reflection and diffuse reflection features from the angle information of reflected light on the skin surface by filtering scattered light data, thereby achieving effective analysis and feature extraction of reflected light from the skin surface. Existing technologies typically use raw reflected light data directly, but this often contains a large amount of background noise or irrelevant reflection information, resulting in poor performance when extracting accurate features. Therefore, this invention reduces interference from irrelevant scattered light by presetting a specific reflection direction when acquiring the reflected light image. First, scattered light data from the skin surface is acquired using a sensor or imaging device. Then, this data is filtered according to the preset direction to extract reflected light intensity information at different angles. The techniques used in this step may include ray tracing or projection methods in image processing algorithms to achieve accurate acquisition of light intensity images. The final generated reflected light image shows the light intensity distribution on the skin surface at different angles. By selecting a preset direction for the reflected light image, the directions of specular and diffuse reflection can be effectively distinguished, providing clearer light intensity information. This helps subsequent steps accurately extract the specular and diffuse reflection features of the skin. Then, based on the angular distribution of reflected light in the reflected light image, algorithms (such as angle clustering algorithms or threshold-based classification methods) are used to extract the angular ranges of specular and diffuse reflection. This method distinguishes reflection angles according to physical principles, ensuring physical consistency for each category's angle set. Accurately distinguishing between specular and diffuse reflection angle sets ensures more targeted analysis of skin surface reflection features, avoiding errors caused by mixing the two types of reflected light. Significant differences exist in the light reflection characteristics of different regions of the skin surface. Specular reflection is mainly concentrated on smooth surfaces or high-reflectivity areas, while diffuse reflection occurs more frequently on rough or porous surfaces. By analyzing the specular reflection angle set and the diffuse reflection angle set, the spatial distribution of these two types of reflected light can be further determined. Using the obtained specular reflection angle set and diffuse reflection angle set, spatial mapping algorithms (such as 3D reconstruction or region partitioning methods) are employed to spatially locate and extract the specular reflection and diffuse reflection regions. Accurate acquisition of the spatial distribution information of specular and diffuse reflection allows for more precise modeling and analysis of the optical properties of the skin surface, avoiding feature extraction errors caused by inaccurate spatial distribution. Skin specular reflection features are one of the important indicators for analyzing the optical properties of the skin surface, and are usually closely related to the smoothness and moisture content of the skin surface. Based on the spatial distribution of the specular reflection region, optical analysis techniques (such as reflectivity acquisition or surface smoothness analysis) are used to extract the skin specular reflection features. Through comprehensive processing of reflected light intensity and angle information, the optical characteristics of the region are extracted. Meanwhile, skin diffuse reflection features are important parameters for analyzing skin surface roughness and epidermal structure.By analyzing the spatial distribution of diffuse reflection regions, we can further extract the diffuse reflection features of the skin, helping to understand its physical properties, such as roughness and texture. Based on the spatial distribution of diffuse reflection regions, an algorithm similar to specular reflection feature extraction is used to analyze the reflected light intensity, texture features, and other physical properties of the region, thereby obtaining the diffuse reflection features of the skin. This information reflects the roughness, texture, and other biophysical characteristics of the skin surface. By extracting the diffuse reflection features of the skin, accurate data can be provided for the analysis of skin texture and roughness.

[0030] In one embodiment, the step of generating skin reflectance data based on the skin specular reflection features and the skin diffuse reflection features includes: S406. Obtain specular reflection and diffuse reflection components based on the skin specular reflection characteristics and skin diffuse reflection characteristics, and obtain the reflectance ratio based on the specular reflection and diffuse reflection components; S407. Obtain the reflectance of the micro-convex structure and the reflectance of the texture boundary of the skin surface based on the reflectance ratio; S408. Obtain skin texture sharpness based on the reflectivity of the micro-convex structure and the reflectivity of the texture boundary; S409. Obtain a synthesized skin feature image based on the skin texture sharpness and the real-time skin image; S410. Generate skin reflectance data by synthesizing an image based on the skin features.

[0031] As described in steps S406-S410 above, this invention generates skin reflected light data by combining skin specular reflection and diffuse reflection characteristics. The light reflection characteristics of skin mainly consist of specular reflection and diffuse reflection, where specular reflection reflects the smoothness of the skin surface, and diffuse reflection reflects the roughness of the skin surface. By analyzing specular and diffuse reflection, the microstructure of the skin surface and its reflection behavior can be accurately captured, thereby inferring the true visual effect of the skin. Existing technologies typically infer the skin's reflection characteristics by simplifying models, which often fails to accurately reproduce the influence of the skin surface microstructure on reflected light. By simultaneously combining specular and diffuse reflection characteristics, and modeling the birefringence characteristics of the skin's micro-convex structure and texture boundaries, skin reflected light can be simulated more accurately. The characteristics of skin reflected light are influenced by multiple factors, among which specular reflection and diffuse reflection are the two most significant forms of reflection. Specular reflection mainly depends on the smoothness of the skin surface, while diffuse reflection is determined by the skin's micro-protrusions and surface irregularities. By acquiring these two reflective components separately, we can gain a clearer understanding of the light reflection characteristics of the skin surface, especially its reflection behavior under complex lighting conditions. Traditional techniques often neglect the interaction between specular and diffuse reflection, or use only a simplified single reflection model. By simultaneously acquiring the ratio of their reflectance, this step can comprehensively consider the proportion of the two reflection modes, obtaining more accurate reflection data. The skin surface is not perfectly smooth but consists of tiny protrusions and texture boundaries. These subtle structures affect the light reflection pattern. By performing high-resolution optical imaging on local areas of the skin surface, we can identify these micro-protrusions and texture boundaries. Then, based on the acquired specular and diffuse reflection data, we obtain the reflectance of the micro-protrusions and texture boundaries, respectively. The skin surface exhibits a certain degree of birefringence, meaning that light rays are deflected in direction when passing through the skin surface. Birefringence is usually caused by the skin's tissue structure (such as the arrangement of collagen and elastic fibers). Analyzing the birefringence properties of the skin surface can further enhance the realism of the skin reflection light model. By analyzing the optical properties of skin tissue (such as refractive index) and the propagation behavior of light on the skin surface, birefringence features are extracted. Then, by acquiring the birefringence properties of the skin surface and real-time skin images, the real light reflection behavior of the skin can be comprehensively considered to generate a skin feature synthesis image. Through high-precision image synthesis technology, the birefringence properties and real-time skin image data are fused to generate a dynamic skin reflection light feature image. This image contains the microstructure, texture features, and light propagation and refraction of the skin surface, which can effectively enhance the realism of the skin reflection light simulation. Finally, skin reflection light data is generated from the skin feature synthesis image. This data contains detailed information reflecting the light reflection characteristics of the skin surface, such as light intensity, directionality, and its impact on the skin. By integrating the data from all previous steps, the final skin reflection light data is generated.This data, processed through advanced optical simulations and algorithms, accurately reflects the light reflection characteristics of the skin under different conditions.

[0032] In one embodiment, the step of obtaining skin state information based on the real-time skin modal feature data and skin reflectance data, and obtaining skin defect information based on the skin state information, includes: S501. Obtain pigment skin concentration, texture roughness, and skin state depth based on the real-time skin modality feature data; S502. Obtain the texture reflectance based on the skin reflectance data; S503. Obtain skin condition information based on the pigment skin concentration, texture roughness, skin condition depth, and texture reflectivity; S504. Obtain multiple skin segmentation regions based on the skin state information; S505. Obtain a skin condition assessment based on each skin segmentation region, and determine whether the skin condition assessment is less than a preset threshold. If it is less than 1 / 3, the skin segmentation area is judged to be in a healthy state. If it is greater than or equal to, the skin segmentation area is judged to be in an unhealthy state; S506. Obtain skin defect information based on the segmented areas of unhealthy skin.

[0033] As described in steps S501-S506 above, this invention acquires skin condition information through real-time skin modal feature data and skin reflective light data, thereby accurately assessing skin health and further analyzing skin defects. Skin modal feature data refers to specific data describing skin condition by collecting information on changes in the skin surface, including pigment concentration and texture roughness. Pigment concentration reflects skin color changes, texture roughness reflects the fineness of the skin surface structure, and skin condition depth reflects the overall health depth of the skin. First, through real-time data acquisition of skin modal features, this step combines sensor or imaging technology to obtain three main parameters: pigment concentration, texture roughness, and skin condition depth. These parameters can assess the surface and deep state of the skin from different perspectives. For example, pigment concentration reflects pigmentation or uneven skin tone, while texture roughness reveals whether the skin is aging or otherwise damaged. Skin reflective light data provides detailed information on the interaction between the skin surface and light. Different skin textures reflect light differently, and changes in texture reflectivity can help identify the microscopic features of the skin. Traditional skin detection methods often fail to fully utilize the interaction between skin and light. This invention, however, enhances the accuracy of skin assessment by meticulously analyzing the skin's reflectivity. Subsequently, an optical sensor illuminates the skin surface to obtain the intensity and distribution of reflected light, further acquiring texture reflectivity. This reflectivity data effectively supplements information about skin surface texture. For example, smooth and healthy skin reflects light more strongly, while skin with scars or signs of aging may exhibit lower reflectivity. Furthermore, skin health is not determined by a single factor but by a complex interplay of multiple factors. Traditional detection methods often overlook the combined influence of factors such as pigment concentration, texture roughness, and reflectivity. By comprehensively analyzing the data obtained in steps S501 and S502 (pigment concentration, texture roughness, skin condition depth, and texture reflectivity), a comprehensive skin condition information is formed. For instance, combining pigment concentration and texture roughness allows for a more accurate assessment of the presence of age spots or signs of aging, while reflectivity analysis reveals whether the skin surface is damaged. Different areas of the skin may exhibit varying health conditions. Traditional methods often assess the skin as a whole, neglecting the differences between localized areas. This method divides the skin into multiple segments based on a comprehensive assessment of its condition. The skin condition of each segment can be analyzed independently, avoiding the mixing of healthy and unhealthy areas. This step allows for more precise subsequent condition assessment and defect detection for different areas. Since skin health exhibits a gradual change, threshold judgment methods can effectively distinguish between healthy and unhealthy states. Based on the skin condition information of the aforementioned segmented areas, each area is assessed. Then, by comparing this assessment with a preset health threshold, it is determined whether each area falls below the healthy standard.If the evaluation value of a certain area is less than a threshold, it is judged to be in a healthy state; otherwise, it is judged to be in an unhealthy state. Finally, for the skin segments judged to be unhealthy, this step further extracts the skin defect information of that area. For example, if a certain area is identified as unhealthy, the system can detect whether there are spots, wrinkles, scars, or other lesion features in that area, thereby providing detailed skin defect information.

[0034] In one embodiment, the step of generating a skin condition assessment report based on the skin defect information includes: S601. Obtain the defect pigment ratio, defect area ratio, and defect severity based on the skin defect information; S602. Obtain the defect index based on the defect pigment ratio, defect area ratio, and severity. S603. Based on the defect index, a skin condition assessment report is generated according to the defect pigment ratio, defect area ratio, and defect severity.

[0035] As described in steps S601-S603 above, this invention first obtains three key parameters based on skin defect information: defect pigment ratio, defect area ratio, and defect severity. The defect pigment ratio includes: spots, wrinkles, scars, and hyperpigmentation; the defect area ratio represents the proportion of each defect in the entire skin area; the defect severity is assessed by evaluating factors such as the depth, size, and color of the defect. Image data of the skin is acquired through image recognition technology, and image processing algorithms are used to extract skin defect features, accurately identifying the defect pigment ratio and obtaining the defect area ratio. Based on the physical characteristics of the defect, such as color changes, depth, and size, the severity of the defect is evaluated. Subsequently, the obtained defect pigment ratio, defect area ratio, and defect severity are used to generate a defect index. The defect index is a comprehensive indicator, obtained by weighting the defect pigment ratio, defect area ratio, and severity to derive a comprehensive numerical score. This index reflects the overall state of skin health, quantifies the impact of skin defects, and is a key indicator for assessing skin health. By weighting the various data points, skin condition can be represented as a comprehensive index. A first weighting coefficient is obtained based on the defect pigmentation ratio, a second weighting coefficient is obtained based on the defect area ratio, and a third weighting coefficient is obtained based on the severity. The formula for calculating the defect index is as follows: ; Here, represents the defect index, represents the defect pigment ratio, represents the first weighting coefficient, represents the defect area percentage, represents the second weighting coefficient, represents the severity, and represents the third weighting coefficient. Combining the aforementioned information on defect type, defect area percentage, and defect severity, a detailed skin condition assessment report is generated. The purpose of generating this report is to provide users with a clear understanding of their skin condition by summarizing relevant information on skin defects, enabling them to take targeted care or treatment measures. This is achieved by integrating the defect index and other relevant data (such as defect type, defect area percentage, severity, etc.) into a structured report format and presenting it to the user. The report may include the following: the distribution of defect types, the defect area percentage, the severity score for each type of defect, and the comprehensive impact index of the defects. Example 2

[0036] Embodiment 2 of the present invention provides a computer-readable storage medium having a program stored thereon. When the program is executed by a processor, it implements the skin visual detection method based on LED lights provided in Embodiment 1.

[0037] The readable storage medium may be more specifically adopted, including but not limited to: portable disk, hard disk, random access memory, read-only memory, erasable programmable read-only memory, optical storage device, magnetic storage device, or any suitable combination thereof.

[0038] In a possible implementation, the present invention can also be implemented as a program product comprising program code, which, when the program product is run on a terminal device, causes the terminal device to perform the steps of implementing the skin visual detection method based on LED lights in Embodiment 1.

[0039] The program code for executing the present invention can be written in any combination of one or more programming languages. The program code can be executed entirely on the user device, partially on the user device, as a standalone software package, partially on the user device and partially on a remote device, or entirely on a remote device.

[0040] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in this application and in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual-speed SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0041] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0042] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent results or equivalent process transformations made based on the content of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A skin visual detection method based on LED lights, characterized in that, include: Acquire a multi-band LED light source and generate multi-band detection light based on the multi-band LED light source; Based on the multi-band detection light, the skin's reflectance spectrum data and filtered scattered light data are acquired, and the image information to be identified is generated based on the reflectance spectrum data; Based on the image information to be identified, obtain skin pigment information on the skin surface, and generate real-time skin modality feature data based on the skin pigment information; Based on the filtered scattered light data, obtain the skin specular reflection characteristics and skin diffuse reflection characteristics of the skin surface, and generate skin reflected light data based on the skin specular reflection characteristics and skin diffuse reflection characteristics; Skin condition information is obtained based on the real-time skin modal feature data and skin reflectance data, and skin defect information is obtained based on the skin condition information; A skin condition assessment report is generated based on the skin defect information.

2. The skin visual detection method based on LED lights according to claim 1, characterized in that, The step of generating multi-band detection light based on the multi-band LED light source includes: The multi-band LED light source generates a multi-spectral illumination band, wherein the multi-spectral illumination band sequentially activates the ultraviolet, blue, green, and red light bands; An angular diffusion process is performed on the multispectral illumination band, and the multispectral diffused illumination band after diffusion processing is obtained. Obtain the ambient light intensity in real time; Based on the established mapping relationship between multispectral diffused illumination bands and ambient light intensity, the illumination-environment mapping correlation degree is obtained; The multi-band LED light source is corrected based on the lighting-environment mapping correlation to obtain the multi-band detection light.

3. The skin visual detection method based on LED lights according to claim 1, characterized in that, The step of generating image information to be identified based on the reflectance spectral data includes: Based on the reflected spectral data, multi-band reflected spectral data of skin reflection is obtained, wherein the multi-band reflected spectral data includes UV light signals and visible light reflected signals; Based on the multi-band reflectance spectral data, UV simulation images and reflected light simulation images are generated; Obtain the reflection position coordinate information of the simulated reflected light image; Obtain the real-time skin image corresponding to the reflection position coordinates; The generated UV simulation image, reflected light simulation image, and real-time skin image are aligned, and skin feature information is extracted; The image information to be identified is generated based on the skin feature information.

4. The skin visual detection method based on LED lights according to claim 1, characterized in that, The step of obtaining skin pigment information on the skin surface based on the image information to be identified includes: Pixel feature point information is extracted based on the image information to be identified; The skin reflectance ratio is obtained from the simulated image of reflected light. A melanin concentration-reflectance mapping relationship is established based on the pixel feature point information and the skin reflectance ratio. The melanin concentration content is obtained based on the melanin concentration-reflectance mapping relationship. The real-time contrast between the melanin deposition area and normal skin is obtained based on the pixel feature point information. A pigment index distribution map is generated based on the real-time contrast. Skin pigmentation information on the skin surface is obtained based on the pigment index distribution map.

5. The skin visual detection method based on LED lights according to claim 1, characterized in that, The step of generating real-time skin modality feature data based on skin pigmentation information includes: The location information of raised areas and recessed areas on the skin surface is obtained based on the real-time skin image. The skin surface roughness is obtained based on the location information of the protruding and concave areas. Based on the skin pigment information, obtain pigment grading texture information, and extract skin texture features based on the pigment grading texture information; A roughness-texture correlation mapping is established based on the skin surface roughness and skin texture features, and the roughness-texture correlation degree is obtained; Skin elasticity characteristics are obtained based on the roughness-texture correlation. Real-time skin modal feature data is generated based on the skin elasticity characteristics.

6. The skin visual detection method based on LED lights according to claim 1, characterized in that, The step of obtaining the specular reflection features and diffuse reflection features of the skin surface based on the filtered scattered light data includes: A reflected light image in a preset direction is obtained based on the filtered scattered light data; The specular reflection angle set and diffuse reflection angle set of the skin surface are obtained based on the reflected light image; The spatial distribution of the specular reflection region and the spatial distribution of the diffuse reflection region are obtained based on the specular reflection angle set and the diffuse reflection angle set. The skin's specular reflection characteristics are obtained based on the spatial distribution of the specular reflection areas. The diffuse reflection characteristics of the skin are obtained based on the spatial distribution of the diffuse reflection region.

7. The skin visual detection method based on LED lights according to claim 1, characterized in that, The step of generating skin reflectance data based on the skin specular reflection characteristics and skin diffuse reflection characteristics includes: The specular and diffuse reflection components are obtained based on the skin specular and diffuse reflection characteristics, and the reflectance ratio is obtained based on the specular and diffuse reflection components. The reflectance of the micro-convex structure and the reflectance of the texture boundary on the skin surface are obtained based on the reflectance ratio. Skin texture sharpness is obtained based on the reflectivity of the micro-convex structure and the reflectivity of the texture boundary. A skin feature synthesis image is obtained based on the skin texture sharpness and the real-time skin image; Skin reflectance data is generated from the synthesized image based on the skin features.

8. The skin visual detection method based on LED lights according to claim 1, characterized in that, The steps of obtaining skin state information based on the real-time skin modal feature data and skin reflectance data, and obtaining skin defect information based on the skin state information, include: Based on the real-time skin modality feature data, pigment concentration, texture roughness, and skin condition depth are obtained; The texture reflectance is obtained based on the skin reflectance data; Skin condition information is obtained based on the skin pigment concentration, texture roughness, skin condition depth, and texture reflectivity. Multiple skin segmentation regions are obtained based on the skin condition information; A skin condition assessment is obtained for each skin segmentation region, and it is determined whether the skin condition assessment is less than a preset threshold. If it is less than 1 / 3, the skin segmentation area is judged to be in a healthy state. If it is greater than or equal to, the skin segmentation area is judged to be in an unhealthy state; Information on skin defects is obtained by segmenting areas of unhealthy skin.

9. The skin visual detection method based on LED lights according to claim 1, characterized in that, The step of generating a skin condition assessment report based on the skin defect information includes: Based on the skin defect information, obtain the defect pigment ratio, defect area ratio, and defect severity; The defect index is obtained based on the defect pigment ratio, defect area ratio, and severity. Based on the defect index, a skin condition assessment report is generated according to the defect pigment ratio, defect area ratio, and defect severity.

10. A computer-readable storage medium, characterized in that, It stores a program that, when executed by a processor, implements the skin visual detection method based on LED lights as described in any one of claims 1-9.

Images