Skin detection system based on multi-angle image acquisition

Abstract

The invention discloses a skin detection system based on multi-angle image acquisition, and relates to the technical field of skin detection.The skin detection system is characterized in that an image acquisition module cooperates with multiple cameras for synchronous shooting to obtain skin images at all angles, an image processing and analyzing module preprocesses the images and extracts feature parameters such as texture and glossiness, and therefore the skin type is judged; the skin color is compared with a standard skin color card to analyze the skin color, the image acquisition control module automatically adjusts the brightness, angle and color temperature of the auxiliary light source and camera shooting parameters by applying sensor feedback and formula calculation according to the skin type and the skin color, and the skin detection module normalizes the feature parameters to form a comprehensive feature vector; the evaluation model is input to obtain the skin quality, the user interaction module displays the skin quality and problems, the skin quality fluctuation coefficient is calculated, the change trend of the skin quality fluctuation coefficient is analyzed, the system achieves accurate detection and analysis of the skin in an all-around mode, and powerful support is provided for personalized skin care.

Description

Technical Field

[0001] The present invention relates to the technical field of skin detection, and particularly to a skin detection system based on multi-angle image acquisition. Background Art

[0002] With the continuous improvement of people's attention to skin health and beauty, skin detection technology has been increasingly valued. Traditional skin detection methods are mostly manual visual inspection or simple single-angle image acquisition and analysis, which have problems such as incomplete detection and low accuracy. Manual visual inspection highly depends on the experience of the inspectors, is highly subjective, and is prone to missing subtle skin problems. Moreover, single-angle image acquisition cannot obtain all-round information of the skin, and it is difficult to accurately detect and evaluate some skin problems with complex surface features, such as lesions hidden in skin folds and skin spots showing different characteristics under different-angle illuminations. Therefore, it is particularly important to invent a skin detection system based on multi-angle image acquisition.

[0003] An existing technology, such as a skin detection system based on multi-angle image acquisition disclosed in the invention patent application with the publication number of CN113393436A, includes: an image acquisition area, a multi-angle image acquisition module, and an image analysis module; wherein, the multi-angle image acquisition module includes a horizontal arc bracket and a vertical arc bracket arranged opposite to the image acquisition area, and at least two image acquisition units evenly distributed along the arc are respectively arranged on the horizontal arc bracket and the vertical arc bracket. The image acquisition units are used to acquire skin images of the detection target in the image acquisition area and transmit the acquired skin images to the image analysis module; the image analysis module is used to perform skin detection processing on the received skin images acquired by each image acquisition unit to obtain skin detection results. This invention helps to improve the reliability and effect of skin image acquisition of the detection target, and indirectly improves the quality and effect of skin condition detection based on images.

[0004] The existing technology also has the following defects, specifically manifested in: 1. In the existing technology, data acquisition has limitations, usually only focusing on a few features of the skin, unable to comprehensively reflect the true state of the skin. The evaluation of skin quality is mostly qualitative description, and the result is not precise enough. The detection results provided to users may only be simple descriptions of skin types or quality, lacking detailed analysis of skin problems and display of skin quality change trends, and unable to provide continuous skin health management suggestions for users.

[0005] 2. In the prior art, the attention to automatically adjusting the relevant device parameters in the detection process based on the user's skin type and skin color is not high. It is unable to automatically adjust the brightness, angle, and color temperature of the auxiliary light source. Manually adjusting the auxiliary light source and camera shooting parameters not only takes time and effort but also makes it difficult to ensure the accuracy and consistency of each adjustment, affecting the reliability of the detection results, increasing the operation steps and time costs. Especially when detecting a large number of users, it will seriously affect the detection efficiency. Summary of the Invention

[0006] The purpose of the present invention is to provide a skin detection system based on multi-angle image acquisition, which solves the problems existing in the background technology.

[0007] To solve the above technical problems, the present invention adopts the following technical solutions: The present invention provides a skin detection system based on multi-angle image acquisition, including: an image acquisition module, which is used to coordinate the synchronous shooting of multiple cameras and obtain the skin images of the user at each angle at the same moment.

[0008] An image processing and analysis module, which is used to preprocess the skin images of the user at each angle, extract the characteristic parameters of the skin of the user at each angle by using a variety of image processing algorithms, determine the user's skin type, and analyze the user's skin color.

[0009] An image acquisition control module, which is used to automatically adjust the relevant parameters of the auxiliary light source according to different skin types through an intelligent control system, and automatically adjust the shooting parameters of the camera based on the user's skin type and skin color.

[0010] A skin detection module, which evaluates the skin quality of the user and analyzes the changing trend of the user's skin quality based on the characteristic parameters of the skin of the user at each angle.

[0011] A user interaction module, which is used to display the user's skin quality and the changing trend of the user's skin quality.

[0012] Preferably, the method for determining the user's skin type is as follows: Extract the characteristic parameters of the user's skin from various angles. The characteristic parameters include roughness, surface reflectivity, sebum secretion amount, and pore diameter. Compare them with the characteristic parameters corresponding to each skin type stored in the database. If the characteristic parameters of the user's skin from various angles are consistent with the characteristic parameters corresponding to oily skin stored in the database, it is determined that the user's skin type is oily. If the characteristic parameters of the user's skin from various angles are consistent with the characteristic parameters corresponding to dry skin stored in the database, it is determined that the user's skin type is dry. If the characteristic parameters of the user's skin from various angles are consistent with the characteristic parameters corresponding to neutral skin stored in the database, it is determined that the user's skin type is neutral. If the characteristic parameters of the user's skin from various angles are consistent with the characteristic parameters corresponding to combination skin stored in the database, it is determined that the user's skin type is combination. If the characteristic parameters of the user's skin from various angles are consistent with the characteristic parameters corresponding to sensitive skin stored in the database, it is determined that the user's skin type is sensitive.

[0013] Based on this judgment, the user's skin type is divided into oily, dry, neutral, combination, and sensitive.

[0014] Preferably, the method for analyzing the user's skin color is as follows: Extract the skin images of the user from various angles, and use a fusion algorithm to obtain the user's skin image. Compare it with the standard skin color card images stored in the database respectively, and use an image matching algorithm to calculate the color similarity between the user's skin image and each standard skin color card image. Compare them with each other to obtain the standard skin color card image corresponding to the maximum color similarity, and use the skin color corresponding to this standard skin color card image as the user's skin color.

[0015] Preferably, the formula for calculating the color similarity between the user's skin image and each standard skin color card image is: where α m represents the color similarity between the user's skin image and the m-th standard skin color card image, b i represents the RGB value of the i-th pixel point in the user's skin image, b m represents the RGB value of the pixel point of the m-th standard skin color card image, c i represents the HSV value of the i-th pixel point in the user's skin image, c m represents the HSV value of the pixel point of the m-th standard skin color card image, i represents the number of each pixel point, i = 1, 2,..., j, j is a positive integer greater than 2, j represents the number of pixel points, m represents the number of each standard skin color card image, m = 1, 2,..., n, n is a positive integer greater than 2, e represents the natural constant, and φ1 and φ2 respectively represent the appropriate RGB value weight factor and the appropriate HSV value weight factor stored in the database.

[0016] Preferably, the method for specifically analyzing the automatic adjustment of the relevant parameters of the auxiliary light source is as follows: The relevant parameters of the auxiliary light source specifically include brightness, angle, and color temperature.

[0017] Extract the user's skin type, compare it with the target brightness values corresponding to each skin type stored in the database to obtain the target brightness value corresponding to the user's skin type, obtain the ambient light intensity value through a photosensitive sensor, compare the target brightness value corresponding to the user's skin type with the ambient light intensity value, and calculate the brightness deviation value. The calculation formula is: ΔL = L 目标 -L 环境 , where ΔL represents the brightness deviation value, L 目标 represents the target brightness value corresponding to the user's skin type, and L 环境 represents the ambient light intensity value, and then calculate the required light source brightness value. The calculation formula is: where L 所需 represents the required light source brightness value, K p represents the proportionality coefficient stored in the database, K g represents the integral coefficient stored in the database, and K d represents the differential coefficient stored in the database, and then adjust the light source brightness value.

[0018] Based on the user's skin type, compare it with the light source angle values corresponding to each skin type stored in the database to obtain the target angle value corresponding to the user's skin type, obtain the current light source angle value through an angle sensor, compare the target angle value corresponding to the user's skin type with the current light source angle value, calculate the angle deviation value, calculate the required light source angle value, and then adjust the light source angle value.

[0019] Based on the user's skin type, compare it with the color temperature values corresponding to each skin type stored in the database to obtain the target color temperature value corresponding to the user's skin type, obtain the current light color temperature value through a color temperature sensor, compare the target color temperature value corresponding to the user's skin type with the current light color temperature value, calculate the color temperature deviation value, calculate the required light color temperature value, and then adjust the light color temperature value.

[0020] Preferably, the method for specifically analyzing the automatic adjustment of the camera shooting parameters based on the user's skin type and skin color is as follows: Obtain the user's skin type and skin color, extract the characteristic parameters of the user's skin type and skin color, perform normalization processing to obtain the skin type value and skin color type value, input them into the exposure duration model, and the output result is the exposure duration. The expression of the exposure duration model is: where T represents the exposure duration corresponding to the user's skin type and skin color, and T baserepresents the base exposure duration, S represents the skin type value after normalization processing, C represents the skin color type value after normalization processing, and respectively represent the appropriate skin type value weight factor and the appropriate skin color type value weight factor in the exposure duration model stored in the database.

[0021] Obtain the user's skin type and skin color, extract the characteristic parameters of the user's skin type and skin color, perform normalization processing to obtain the skin type value and the skin color type value, and input them into the sensitivity model. The output result is the sensitivity corresponding to the user's skin type and skin color. The sensitivity model expression is: where ISO represents the sensitivity corresponding to the user's skin type and skin color, ISO base represents the base sensitivity, and respectively represent the appropriate skin type value weight factor and the appropriate skin color type value weight factor in the sensitivity model stored in the database.

[0022] Preferably, the method for evaluating the user's skin quality is as follows: Extract the roughness, surface reflectivity, oil secretion amount, and pore diameter of the user's skin, perform normalization processing, and input the processed data into the skin quality model. The output result is the user's skin quality. If β = A, it means the user's skin quality is excellent. If β = B, it means the user's skin quality is good. If β = G, it means the user's skin quality is medium. If β = H, it means the user's skin quality is poor.

[0023] Preferably, the model expression is: where y represents the user's skin quality coefficient, R, S, T, and V respectively represent the roughness, surface reflectivity, oil secretion amount, and pore diameter of the user's skin after normalization processing, γ1, γ2, γ3, and γ4 respectively represent the appropriate roughness weight factor, appropriate surface reflectivity weight factor, appropriate oil secretion amount weight factor, and appropriate pore diameter weight factor stored in the database, A represents the skin quality coefficient range corresponding to excellent skin quality, B represents the skin quality coefficient range corresponding to good skin quality, G represents the skin quality coefficient range corresponding to medium skin quality, and H represents the skin quality coefficient range corresponding to poor skin quality.

[0024] Preferably, for analyzing the changing trend of the user's skin quality, the specific analysis method is as follows: Extract the user's skin quality coefficients of each historical detection from the database, calculate the user's skin quality fluctuation coefficient, and compare it with the appropriate user's skin quality fluctuation coefficient stored in the database. If the user's skin quality fluctuation coefficient is higher than the appropriate user's skin quality fluctuation coefficient stored in the database, it indicates that the user's skin quality shows an upward trend. If the user's skin quality fluctuation coefficient is lower than the appropriate user's skin quality fluctuation coefficient stored in the database, it indicates that the user's skin quality shows a downward trend. If the user's skin quality fluctuation coefficient is equal to the appropriate user's skin quality fluctuation coefficient stored in the database, it indicates that the user's skin quality has not changed.

[0025] Based on this judgment, the changing trend of the user's skin quality is upward, downward, or unchanged.

[0026] Preferably, for calculating the user's skin quality fluctuation coefficient, the calculation formula is: where χ represents the user's skin quality fluctuation coefficient, and y r represents the user's skin quality coefficient of the r-th historical detection, and y r+1 represents the user's skin quality coefficient of the (r + 1)-th historical detection. r represents the number of each historical detection, r = 1, 2,..., s, where s is a positive integer greater than 2, and s represents the number of historical detections.

[0027] The beneficial effects of the present invention are as follows: 1. In the present invention, multiple cameras are used to synchronously collect skin images from multiple angles, and skin types are judged and skin quality is evaluated by comprehensively considering various characteristic parameters such as texture, glossiness, sebum secretion amount, and pore diameter, improving the accuracy of detection. By calculating the skin quality fluctuation coefficient and analyzing the changing trend of skin quality in historical detection data, a basis for long-term skin health management of users is provided.

[0028] 2. In the present invention, more attention is paid to automatically adjusting the relevant device parameters during the detection process based on the user's skin type and skin color. The brightness, angle, and color temperature of the auxiliary light source can be automatically adjusted according to the skin type, and the camera shooting parameters can also be automatically adjusted based on the skin type and skin color, reducing the errors caused by manual intervention, improving the reliability of the detection results, and reducing the operation steps and time costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0030] Figure 1 This is a schematic diagram of the system structure connection of the present invention. Detailed implementation manners

[0031] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0032] Referring to Figure 1 As shown, the present invention provides a skin detection system based on multi-angle image acquisition, including: an image acquisition module for coordinating the synchronous shooting of multiple cameras to obtain skin images of a user from various angles at the same moment.

[0033] An image processing and analysis module for preprocessing the skin images of the user from various angles, using a variety of image processing algorithms to extract the characteristic parameters of the skin of the user from various angles, judging the skin type of the user, and analyzing the skin color of the user.

[0034] In a specific embodiment, the method for specifically analyzing the judgment of the skin type of the user is: extracting the characteristic parameters of the skin of the user from various angles, where the characteristic parameters include: roughness, surface reflectivity, oil secretion amount, and pore diameter, and comparing them with the characteristic parameters corresponding to each skin type stored in the database respectively. If the characteristic parameters of the skin of the user from various angles are consistent with the characteristic parameters corresponding to the oily skin stored in the database, it is determined that the skin type of the user is oily. If the characteristic parameters of the skin of the user from various angles are consistent with the characteristic parameters corresponding to the dry skin stored in the database, it is determined that the skin type of the user is dry. If the characteristic parameters of the skin of the user from various angles are consistent with the characteristic parameters corresponding to the neutral skin stored in the database, it is determined that the skin type of the user is neutral. If the characteristic parameters of the skin of the user from various angles are consistent with the characteristic parameters corresponding to the combination skin stored in the database, it is determined that the skin type of the user is combination. If the characteristic parameters of the skin of the user from various angles are consistent with the characteristic parameters corresponding to the sensitive skin stored in the database, it is determined that the skin type of the user is sensitive.

[0035] Based on this judgment, the skin types of users are divided into oily, dry, neutral, combination, and sensitive.

[0036] In the present invention, skin images are collected from multiple angles synchronously by multiple cameras, and the skin type is judged and the skin quality is evaluated by comprehensively considering various characteristic parameters such as texture, glossiness, oil secretion amount, and pore diameter, improving the accuracy of detection. By calculating the skin quality fluctuation coefficient and analyzing the skin quality change trend in the historical detection data, a basis for long-term skin health management for users is provided.

[0037] In a specific embodiment, the skin color of the user is analyzed. The specific analysis method is as follows: Skin images of the user from various angles are extracted, and a fusion algorithm is used to obtain the skin image of the user. The obtained skin image is respectively compared with the standard skin color card images stored in the database. An image matching algorithm is used to calculate the color similarity between the skin image of the user and each standard skin color card image. By comparing them with each other, the standard skin color card image corresponding to the maximum color similarity is obtained, and the skin color corresponding to this standard skin color card image is used as the skin color of the user.

[0038] It should be noted that the fusion algorithm adopted is a prior art, which is specifically divided into image preprocessing, image fusion algorithm, and image postprocessing. It can stitch the skin images of the user from various angles into a complete picture to more comprehensively observe and analyze the skin condition of the user.

[0039] In a specific embodiment, the formula for calculating the color similarity between the skin image of the user and each standard skin color card image is as follows: where α m represents the color similarity between the skin image of the user and the m-th standard skin color card image, b i represents the RGB value of the i-th pixel point in the skin image of the user, b m represents the RGB value of the pixel point in the m-th standard skin color card image, c i represents the HSV value of the i-th pixel point in the skin image of the user, c m represents the HSV value of the pixel point in the m-th standard skin color card image, i represents the number of each pixel point, i = 1, 2,..., j, j is a positive integer greater than 2, j represents the number of pixel points, m represents the number of each standard skin color card image, m = 1, 2,..., n, n is a positive integer greater than 2, e represents the natural constant, and φ1 and φ2 respectively represent the appropriate RGB value weight factor and the appropriate HSV value weight factor stored in the database.

[0040] It should be noted that RGB is an additive color model, which represents various colors through different intensity combinations of three primary colors: red (R), green (G), and blue (B). The RGB value calculation formula is: RGB = ζ1*R + ζ2*G + ζ3*B, where ζ1, ζ2, and ζ3 respectively represent the appropriate R value weight factor, the appropriate G value weight factor, and the appropriate B value weight factor, which are set by professionals. HSV, that is, hue, saturation, and value, is a color representation method that is more in line with human visual perception. The calculation method is as above and will not be elaborated here. The appropriate RGB value weight factor and the appropriate HSV value weight factor stored in the database are set by professionals.

[0041] An image acquisition control module is used to automatically adjust the relevant parameters of the auxiliary light source according to different skin types through an intelligent control system, and automatically adjust the shooting parameters of the camera based on the user's skin type and skin color.

[0042] In a specific embodiment, the method for specifically analyzing the automatic adjustment of the relevant parameters of the auxiliary light source is as follows: the relevant parameters of the auxiliary light source specifically include brightness, angle, and color temperature.

[0043] Extract the user's skin type, compare it with the target brightness value corresponding to each skin type stored in the database to obtain the target brightness value corresponding to the user's skin type, obtain the ambient light intensity value through a photosensitive sensor, compare the target brightness value corresponding to the user's skin type with the ambient light intensity value, and calculate the brightness deviation value. The calculation formula is: ΔL = L 目标 -L 环境 where ΔL represents the brightness deviation value, L 目标 represents the target brightness value corresponding to the user's skin type, and L 环境 represents the ambient light intensity value, and then calculate the required light source brightness value. The calculation formula is: where L 所需 represents the required light source brightness value, K p represents the proportionality coefficient stored in the database, K g represents the integral coefficient stored in the database, and K d represents the differential coefficient stored in the database, and then adjust the light source brightness value.

[0044] In the present invention, the attention to automatically adjusting the relevant device parameters during the detection process based on the user's skin type and skin color is increased. The brightness, angle, and color temperature of the auxiliary light source can be automatically adjusted according to the skin type, and the camera shooting parameters can also be automatically adjusted based on the skin type and skin color, reducing the errors caused by manual intervention, improving the reliability of the detection results, and reducing the operation steps and time costs.

[0045] It should be noted that the proportionality coefficient is a linear adjustment factor. In the formula for calculating the required light source brightness value, it is directly multiplied by the brightness deviation value. This means that when the brightness deviation value changes, the proportionality coefficient determines the basic amplitude of the light source brightness adjustment. The integral coefficient is mainly used to eliminate the steady-state deviation of the system. During the brightness adjustment process, due to various factors, the light source brightness may not accurately reach the target brightness value, resulting in a persistent small deviation. The integral term will accumulate this persistent deviation and adjust the light source brightness according to the accumulated result. Over time, this steady-state deviation will be gradually eliminated. The differential coefficient is used to predict the change trend of the brightness deviation. It adjusts the light source brightness according to the change rate of the brightness deviation. If the change rate of the brightness deviation is large, it indicates that the ambient light intensity or the target brightness value corresponding to the user's skin type is changing rapidly. The differential coefficient will cause the light source brightness to respond in advance to adapt to this rapid change situation, which is set by professionals.

[0046] Based on the user's skin type, compare it with the light source angle values corresponding to each skin type stored in the database to obtain the target angle value corresponding to the user's skin type. Through the angle sensor, obtain the current light source angle value. Compare the target angle value corresponding to the user's skin type with the current light source angle value, calculate the angle deviation value, calculate the required light source angle value, and then adjust the light source angle value.

[0047] Based on the user's skin type, compare it with the color temperature values corresponding to each skin type stored in the database to obtain the target color temperature value corresponding to the user's skin type. Through the color temperature sensor, obtain the current light color temperature value. Compare the target color temperature value corresponding to the user's skin type with the current light color temperature value, calculate the color temperature deviation value, calculate the required light color temperature value, and then adjust the light color temperature value.

[0048] In a specific embodiment, the method for automatically adjusting the shooting parameters of the camera based on the user's skin type and skin color is as follows: Obtain the user's skin type and skin color, extract the characteristic parameters of the user's skin type and skin color, perform normalization processing to obtain the skin type value and skin color type value, and input them into the exposure duration model. The output result is the exposure duration. The expression of the exposure duration model is: where T represents the exposure duration corresponding to the user's skin type and skin color, T base represents the basic exposure duration, S represents the skin type value after normalization processing, C represents the skin color type value after normalization processing, and respectively represent the appropriate skin type value weight factor and appropriate skin color type value weight factor in the exposure duration model stored in the database.

[0049] Obtain the user's skin type and skin color, extract the characteristic parameters of the user's skin type and skin color, perform normalization processing to obtain the skin type value and skin color type value, and input them into the sensitivity model. The output result is the sensitivity corresponding to the user's skin type and skin color. The expression of the sensitivity model is: where ISO represents the sensitivity corresponding to the user's skin type and skin color, and ISO base represents the base sensitivity, and respectively represent the weight factors of the appropriate skin type value and the weight factor of the appropriate skin color type value in the sensitivity model stored in the database.

[0050] It should be noted that the weight factors of the appropriate skin type value and the weight factor of the appropriate skin color type value in the exposure duration model stored in the database, and the weight factors of the appropriate skin type value and the weight factor of the appropriate skin color type value in the sensitivity model stored in the database are set by professionals.

[0051] The skin detection module evaluates the user's skin quality based on the characteristic parameters of the user's skin from various angles and analyzes the changing trend of the user's skin quality.

[0052] In a specific embodiment, the method for evaluating the user's skin quality is as follows: Extract the roughness, surface reflectivity, oil secretion amount, and pore diameter of the user's skin, perform normalization processing, and input the processed data into the skin quality model. The output result is the user's skin quality. If β = A, it means the user's skin quality is excellent; if β = B, it means the user's skin quality is good; if β = G, it means the user's skin quality is medium; if β = H, it means the user's skin quality is poor.

[0053] In a specific embodiment, the model expression is: where y represents the user's skin quality coefficient, R, S, T, and V respectively represent the roughness, surface reflectivity, oil secretion amount, and pore diameter of the user's skin after normalization processing, γ1, γ2, γ3, and γ4 respectively represent the appropriate roughness weight factor, appropriate surface reflectivity weight factor, appropriate oil secretion amount weight factor, and appropriate pore diameter weight factor stored in the database, A represents the range of skin quality coefficients corresponding to excellent skin quality, B represents the range of skin quality coefficients corresponding to good skin quality, G represents the range of skin quality coefficients corresponding to medium skin quality, and H represents the range of skin quality coefficients corresponding to poor skin quality.

[0054] It should be noted that the appropriate roughness weight factor, appropriate surface reflectance weight factor, appropriate sebum secretion amount weight factor, and appropriate pore diameter weight factor belonging to the village in the database, the skin quality coefficient ranges corresponding to excellent skin quality, good skin quality, medium skin quality, and poor skin quality are set by professionals.

[0055] In a specific embodiment, the method for specifically analyzing the change trend of the user's skin quality is as follows: Extract the user's skin quality coefficients of each historical detection from the database, calculate the user's skin quality fluctuation coefficient, and compare it with the appropriate user's skin quality fluctuation coefficient stored in the database. If the user's skin quality fluctuation coefficient is higher than the appropriate user's skin quality fluctuation coefficient stored in the database, it indicates that the user's skin quality shows an upward trend. If the user's skin quality fluctuation coefficient is lower than the appropriate user's skin quality fluctuation coefficient stored in the database, it indicates that the user's skin quality shows a downward trend. If the user's skin quality fluctuation coefficient is equal to the appropriate user's skin quality fluctuation coefficient stored in the database, it indicates that the user's skin quality has not changed.

[0056] Based on this judgment, the change trend of the user's skin quality is upward, downward, or unchanged.

[0057] It should be noted that the appropriate user's skin quality fluctuation coefficient stored in the database is set by professionals.

[0058] In a specific embodiment, the formula for calculating the user's skin quality fluctuation coefficient is: where χ represents the user's skin quality fluctuation coefficient, and y r represents the user's skin quality coefficient of the rth historical detection, and y r+1 represents the user's skin quality coefficient of the (r + 1)th historical detection. r represents the number of each historical detection, r = 1, 2,..., s, s is a positive integer greater than 2, and s represents the number of historical detections.

[0059] The user interaction module is used to display the user's skin quality and the change trend of the user's skin quality.

[0060] The image acquisition module is connected to the image processing and analysis module, the image processing and analysis module is connected to the image acquisition control module, the image acquisition control module is connected to the skin detection module, the skin detection module is connected to the user interaction module, and the image acquisition control module, the image processing and analysis module, and the skin detection module are simultaneously connected to the database.

[0061] The above content is only an example and illustration of the concept of the present invention. Those skilled in the art of the present technology can make various modifications or supplements to the described specific embodiments or use similar ways to replace them. As long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, they shall fall within the protection scope of the present invention.

Claims

1. A skin detection system based on multi-angle image acquisition, characterized in that, Including: An image acquisition module, which is used to coordinate the synchronous shooting of multiple cameras and obtain the skin images of the user from various angles at the same time. An image processing and analysis module, which is used to preprocess the skin images of the user from various angles, extract the characteristic parameters of the skin of the user from various angles by using a variety of image processing algorithms, judge the skin type of the user, and analyze the skin color of the user. An image acquisition control module, which is used to automatically adjust the relevant parameters of the auxiliary light source according to different skin types through an intelligent control system, and automatically adjust the shooting parameters of the camera based on the skin type and skin color of the user. A skin detection module, which evaluates the skin quality of the user based on the characteristic parameters of the skin of the user from various angles and analyzes the changing trend of the user's skin quality. A user interaction module, which is used to display the user's skin quality and the changing trend of the user's skin quality.

2. The skin detection system based on multi-angle image acquisition according to claim 1, wherein, The specific analysis method for judging the skin type of the user is as follows: Extract the characteristic parameters of the skin of the user from various angles. The characteristic parameters include: roughness, surface reflectivity, oil secretion amount, and pore diameter. Compare them with the characteristic parameters corresponding to each skin type stored in the database respectively. If the characteristic parameters of the skin of the user from various angles are consistent with the characteristic parameters corresponding to the oily skin stored in the database, it is determined that the skin type of the user is oily. If the characteristic parameters of the skin of the user from various angles are consistent with the characteristic parameters corresponding to the dry skin stored in the database, it is determined that the skin type of the user is dry. If the characteristic parameters of the skin of the user from various angles are consistent with the characteristic parameters corresponding to the neutral skin stored in the database, it is determined that the skin type of the user is neutral. If the characteristic parameters of the skin of the user from various angles are consistent with the characteristic parameters corresponding to the combination skin stored in the database, it is determined that the skin type of the user is combination. If the characteristic parameters of the skin of the user from various angles are consistent with the characteristic parameters corresponding to the sensitive skin stored in the database, it is determined that the skin type of the user is sensitive. According to this judgment, the skin types of the user are divided into oily, dry, neutral, combination, and sensitive.

3. The skin detection system based on multi-angle image acquisition according to claim 2, wherein, The specific analysis method for analyzing the skin color of the user is as follows: Extract the skin images of the user from various angles, use a fusion algorithm to obtain the skin image of the user, compare it with the images of each standard skin color card stored in the database respectively, use an image matching algorithm to calculate the color similarity between the skin image of the user and the images of each standard skin color card, compare them with each other, obtain the image of the standard skin color card corresponding to the maximum color similarity, and use the skin color corresponding to the image of the standard skin color card as the skin color of the user.

4. The skin detection system based on multi-angle image acquisition according to claim 3, wherein The color similarity between the user's skin image and each standard skin color card image is calculated by the following formula: where α m represents the color similarity between the user's skin image and the m-th standard skin color card image, b i represents the RGB value of the i-th pixel point in the user's skin image, b m represents the RGB value of the pixel point in the m-th standard skin color card image, c i represents the HSV value of the i-th pixel point in the user's skin image, c m represents the HSV value of the pixel point in the m-th standard skin color card image, i represents the number of each pixel point, i = 1, 2,..., j, j is a positive integer greater than 2, j represents the number of pixel points, m represents the number of each standard skin color card image, m = 1, 2,..., n, n is a positive integer greater than 2, e represents the natural constant, and φ1 and φ2 respectively represent the appropriate RGB value weight factor and the appropriate HSV value weight factor stored in the database.

5. The skin detection system based on multi-angle image acquisition according to claim 1, characterized in that, The specific analysis method for automatically adjusting the relevant parameters of the auxiliary light source is as follows: The relevant parameters of the auxiliary light source specifically include brightness, angle, and color temperature. Extract the user's skin type, compare it with the target brightness value corresponding to each skin type stored in the database to obtain the target brightness value corresponding to the user's skin type. Obtain the ambient light intensity value through a photosensitive sensor, compare the target brightness value corresponding to the user's skin type with the ambient light intensity value, and calculate the brightness deviation value. The calculation formula is: ΔL = L 目标 -L 环境 , where ΔL represents the brightness deviation value, L 目标 represents the target brightness value corresponding to the user's skin type, and L 环境 represents the ambient light intensity value. Then calculate the required light source brightness value. The calculation formula is: where L 所需 represents the required light source brightness value, K p represents the proportionality coefficient stored in the database, K g represents the integral coefficient stored in the database, and K d represents the differential coefficient stored in the database. Then adjust the light source brightness value; Based on the skin type of the user, compare it with the light source angle values corresponding to each skin type stored in the database to obtain the target angle value corresponding to the skin type of the user. Through an angle sensor, obtain the current light source angle value. Compare the target angle value corresponding to the skin type of the user with the current light source angle value, calculate the angle deviation value, calculate the required light source angle value, and then adjust the light source angle value. Based on the user's skin type, compare it with the color temperature values corresponding to each skin type stored in the database to obtain the target color temperature value corresponding to the user's skin type. Through the color temperature sensor, obtain the current light color temperature value. Compare the target color temperature value corresponding to the user's skin type with the current light color temperature value, calculate the color temperature deviation value, calculate the required light color temperature value, and then adjust the light color temperature value.

6. The skin detection system based on multi-angle image acquisition according to claim 1, wherein Automatically adjust the shooting parameters of the camera based on the user's skin type and skin color. The specific analysis method is as follows: Obtain the user's skin type and skin color, extract the characteristic parameters of the user's skin type and skin color, perform normalization processing to obtain the skin type value and skin color type value, and input them into the exposure duration model. The output result is the exposure duration. The expression of the exposure duration model is: where T represents the exposure duration corresponding to the user's skin type and skin color, T base represents the basic exposure duration, S represents the skin type value after normalization processing, C represents the skin color type value after normalization processing, and respectively represent the weight factor of the suitable skin type value and the weight factor of the suitable skin color type value in the exposure duration model stored in the database; Obtain the user's skin type and skin color, extract the characteristic parameters of the user's skin type and skin color, perform normalization processing to obtain the skin type value and skin color type value, and input them into the sensitivity model. The output result is the sensitivity corresponding to the user's skin type and skin color. The expression of the sensitivity model is: where ISO represents the sensitivity corresponding to the user's skin type and skin color, and ISO base represents the base sensitivity, and respectively represent the weight factor of the suitable skin type value and the weight factor of the suitable skin color type value in the sensitivity model stored in the database.

7. The skin detection system based on multi-angle image acquisition according to claim 1, characterized in that Evaluate the user's skin quality. The specific implementation method is as follows: Extract the roughness, surface reflectivity, sebum secretion amount, and pore diameter of the user's skin, perform normalization processing, input the processed data into the skin quality model, and the output result is the user's skin quality. If β = A, it means the user's skin quality is excellent; if β = B, it means the user's skin quality is good; if β = G, it means the user's skin quality is medium; if β = H, it means the user's skin quality is poor.

8. A skin detection system based on multi-angle image acquisition according to claim 7, characterized in that, The model expression is as follows: Where y represents the user's skin quality coefficient, R, S, T, and V respectively represent the roughness, surface reflectivity, oil secretion amount, and pore diameter of the user's skin after normalization, γ1, γ2, γ3, and γ4 respectively represent the appropriate roughness weight factor, appropriate surface reflectivity weight factor, appropriate oil secretion amount weight factor, and appropriate pore diameter weight factor in the database, A represents the skin quality coefficient range corresponding to excellent skin quality, B represents the skin quality coefficient range corresponding to good skin quality, G represents the skin quality coefficient range corresponding to medium skin quality, and H represents the skin quality coefficient range corresponding to poor skin quality.

9. The skin detection system based on multi-angle image acquisition according to claim 8, wherein, Analyze the changing trend of the user's skin quality. The specific analysis method is as follows: Extract the user's skin quality coefficients of each historical detection from the database, calculate the user's skin quality fluctuation coefficient, and compare it with the appropriate user skin quality fluctuation coefficient stored in the database. If the user's skin quality fluctuation coefficient is higher than the appropriate user skin quality fluctuation coefficient stored in the database, it means the user's skin quality shows an upward trend; if the user's skin quality fluctuation coefficient is lower than the appropriate user skin quality fluctuation coefficient stored in the database, it means the user's skin quality shows a downward trend; if the user's skin quality fluctuation coefficient is equal to the appropriate user skin quality fluctuation coefficient stored in the database, it means the user's skin quality has not changed. Based on this judgment, the changing trend of the user's skin quality is upward, downward, or unchanged.

10. The skin detection system based on multi-angle image acquisition according to claim 9, characterized in that, The calculated user skin quality fluctuation coefficient has the following calculation formula: where χ represents the user skin quality fluctuation coefficient, and y r represents the user skin quality coefficient of the r-th historical detection, and y r+1 represents the user skin quality coefficient of the (r + 1)-th historical detection. r represents the number of each historical detection, r = 1, 2,..., s, where s is a positive integer greater than 2 and represents the number of historical detections.

Images