KR20210079090A - Mobile Skin Photography Device and Image Analysis System - Google Patents

Abstract

The present invention is a technology for diagnosing skin diseases on the basis of optics and provide a high-resolution skin diagnostic device in which a plurality of light sources (general visible light, polarized light, and ultraviolet light source) are integrated into the device and applies various algorithms to a skin image captured and enlarged on the basis of the device to analyze conditions, such as sebum, wrinkles, pores, black spots, erythema, blackheads, and dead skin cells and, at the same time, a moisture sensor is included to measure the level of moisture. According to the present invention, since differentiation of skin characteristic patterns is limited only with a simple light source (normal visible light) when capturing skin images, a light source is subdivided for refinement. The skin diagnosis device must be portable and can be control through communication data and applications for interlocking with the mobile environment. After image processing algorithm techniques are applied to the captured images and a definition for analysis is established on the basis of skin troubles and patterns expressed in the images, patterns such as color, shape, thickness, and direction of an object are recognized to quantify the ratio of a predetermined pattern (feature) with respect to an observation range and moisture is measured by measuring the skin condition with a sensor for measuring the impedance of the skin epidermis so that the skin images are analyzed as patterns to classify, match, and categorize the skin image into specific characteristics.

Description

Mobile Skin Photography Device and Image Analysis System

The present invention relates to a customized cosmetic skin diagnosis and skin analysis technology, and more specifically, a pattern (image processing) of an image image optically observed in the diagnostic device by linking an image captured by a diagnostic device using various light sources with a mobile terminal and an app. Result) as a method of classifying, matching, and typology with specific skin characteristics, it further relates to a technology that can produce personalized cosmetics according to the typed analysis result.

In skin diagnosis, a technique for diagnosing one's skin condition through a questionnaire has been introduced and applied, but with the recent development of optical technology, various techniques for image analysis are being developed.

Customized cosmetics can be implemented through the process of diagnosing the user's skin, analyzing the diagnostic data to check the current user's skin condition, and then recommending or manufacturing the best cosmetics for the user's skin condition based on the data. have.

Among these, the most basic starts from how accurately the user's skin condition can be diagnosed, and the most basic part in the development of customized cosmetics is skin diagnosis technology.

In the cosmetic field, there are three types of skin diagnosis: the method of identifying individual skin information through a questionnaire, the method of identifying unique skin characteristics through skin genetic diagnosis, and the method of analyzing data through optical imaging and image processing. .

In order to secure a diagnostic device that recommends customized oil cosmetics based on the skin diagnosis result as a model, it is necessary to develop a skin diagnosis device hardware and an analysis algorithm based on the optical analysis skin diagnosis technology according to the present invention.

KR 1927286 B1 (2018.12.04) YKC Tech Co., Ltd. "UNMANNED SKIN DIAGNOSIS APPARATUS FOR FACE AND BACK OF HAND}" is a kiosk including a touch screen for an unmanned facial and back skin analysis device ; an opening positioned in the kiosk to insert a user's hand; a first skin measurement module located at the top of the kiosk and including a first lighting module for illuminating the user's face and a first camera module for photographing the user's face; a second skin measurement module located in the opening and including a second lighting module for illuminating the user's hand and a second camera module for photographing the user's hand; a control module for calculating a skin diagnosis result of the user based on at least one of a first image image captured by the first skin measurement module and a second image image captured by the second skin measurement module; a storage module configured to store at least one of the first image image and the second image image, an operation result of the control module, and the skin diagnosis result; a print module located in the kiosk and configured to print the operation result of the control module and the skin diagnosis result; and a communication module for transmitting the skin diagnosis result to at least one of a mobile terminal, e-mail, SMS, and SNS through a wired/wireless network, wherein the touch screen operates the unmanned facial and back skin analysis device by touch input, , to display the skin diagnosis result. KR 2015-7030311 (2014.03.24) Shiseido Co., Ltd. "skin darkening evaluation device and skin darkening evaluation method" is a skin darkening evaluation device, frequency analysis means for decomposing into color tone unevenness of, and color tone unevenness for obtaining skin tone information from the color tone unevenness decomposed by the frequency analysis means, and obtaining a tone unevenness index of the color tone unevenness from the acquired skin tone information Disclosed are index acquisition means and darkening evaluation means for evaluating skin darkening corresponding to the skin image based on the color tone unevenness index obtained by the color tone unevenness index acquisition means.

An object of the present invention is to classify, match, and categorize the image captured by the diagnostic device using various light sources by linking the image captured by the diagnostic device with the app of the mobile terminal, which is the analyzer, to analyze the optically observed image image in the diagnostic device as a pattern to classify, match, and typify the specific characteristics of the skin. (Typology) to provide a mobile skin imaging device and image analysis system.

The present invention is a technology for diagnosing skin based on optics. A plurality of light sources (general visible light, polarized light, and ultraviolet light source) are integrated into the device, and various algorithms are applied to the enlarged skin image based on this. It can analyze the condition of pores, black spots, erythema, blackheads, dead skin cells, etc., and provides a high-resolution skin diagnostic device that measures the level of moisture by installing a moisture sensor at the same time. Since there is a limit to the differentiation of characteristic patterns, the light source is subdivided and implemented for product differentiation and elaboration of results. After establishing the definition for analysis based on skin troubles and patterns expressed in images by applying the image processing algorithm technique to ) ratio, and moisture can measure the skin condition by using the impedance measuring sensor of the skin epidermis, and a mobile skin imaging device and image analysis that analyzes the skin image image as a pattern to classify, match, and categorize it into specific characteristics provide the system.

According to the present invention, the skin is diagnosed based on optics and the LED light source (normal visible light, polarized light, ultraviolet light source) is integrated into the device, and various algorithms are applied to the enlarged skin image based on this to apply sebum (oil) and wrinkles. , pores, black spots, erythema, blackheads, dead skin cells, etc. can be analyzed, and a moisture sensor is installed at the same time to measure the moisture level and analyze skin image images as patterns to classify, match, and type specific characteristics. According to the analysis result, there is an effect that customized cosmetics can be manufactured.

1 is a schematic diagram of a 3D model of a portable skin diagnosis device.
2 is a view for explaining a method of using skin observation lighting of a skin diagnosis device.
3 is a diagram illustrating a measurement method and a display example according to a diagnosis item.
4 is a diagram illustrating an example of quantification of skin feature data according to diagnosis items.
5 is a diagram illustrating a H/W circuit diagram of a skin diagnosis device according to an exemplary embodiment.
6A and 6B are diagrams for explaining chip naming of circuit diagram components.
7A to 7D are 3D views of a skin diagnosis device H/W according to an exemplary embodiment.

This technology is a technology for diagnosing skin based on optics. It integrates an LED light source (normal visible light, polarized light, and ultraviolet light source) into the device and applies various algorithms to magnify the skin image based on this to reduce sebum (oil), It is possible to analyze the state of wrinkles, pores, black spots, erythema, blackheads, dead skin cells, etc.

When only a simple light source (general visible light) is used for skin imaging, there is a limit to the differentiation of skin characteristic patterns, so the light source can be subdivided for product differentiation and refinement of results.

In this case, the skin diagnostic device should be portable and should be able to be controlled through data communication and application for interworking with a mobile environment (cell phone, tablet, etc.).

After establishing a definition for analysis based on skin troubles and patterns expressed in images by applying image processing techniques to the captured images, patterns such as color, shape, thickness, and direction of objects are recognized and a specific pattern ( Characteristic) ratio is quantified, and moisture can be utilized by measuring the electrical resistance (impedance) of the skin epidermis.

(1) Customized skin diagnosis technology using artificial intelligence technology

In this study, each evaluation model was developed for moisture, oil, wrinkles, pores, and skin tone, and moisture and oil were divided into four categories (oily, dry, neutral, and combination), and the results of wrinkles, pores, and skin tone were It can be provided by classifying it into grades 1-5.

Based on the results analyzed from the skin diagnosis model, it is applied to a machine learning model to provide an environment for the user to select by outputting a model that recommends cosmetics tailored to individual skin conditions and a list of top three cosmetics suitable for personal skin conditions.

It provides a cloud environment that users can access through design and development of a cloud environment that can be accessed from a cosmetic store or a dermatologist that owns a specific skin diagnosis device.

(2) Facial analysis system that collects and processes skin diagnosis big data and AI-based skin diagnosis algorithm

The purpose of this study is to develop an image-based skin measurement system with data transmission (wired/wireless) function for collecting big data and a skin questionnaire algorithm to assist in skin characteristic analysis.

The image-based skin measurement system can measure pigmentation, wrinkles, pores, and the degree of sunscreen application, and provides images for each skin disease based on artificial intelligence. It optimizes the optimal image modulation conditions for AI-based specific lesion diagnosis, and provides a function that enables treatment linkage through lesion recognition technology linkage.

(3) Development of portable skin diagnosis device and diagnosis management algorithm

This study aims to provide a comprehensive solution that allows users to diagnose and manage their skin based on professional data anytime, anywhere, by linking portable devices and mobile apps.

This portable diagnostic device builds 16 types of analysis DB based on skin type analysis (Baumann skin type) to measure skin, and includes short-range wireless communication between the device and a dedicated app, skin diagnosis algorithm technology, etc.

Skin analysis algorithm uses techniques such as wrinkle detection and quantification algorithm, gray scale imager, contrast enhancement, Otsu threshold, collection of wrinkle information for various age groups, age estimation algorithm, Gaussian blurring, noise tolerance, etc. to detect and quantify skin pores Includes an algorithm, an algorithm that detects and quantitatively analyzes wrinkles and pores in the RGB image of the user's skin, an algorithm that detects and quantitatively analyzes sebum and blemishes in images acquired using ultraviolet rays, and a multi-wavelength skin analysis algorithm for quantifying redness data can do.

On the other hand, as the modern consumption culture enters the era of pursuing one's own taste according to individual personality and values, marketing strategies aimed at 'personalized products' appear throughout the industry. A market has been formed in the form of diagnosing skin conditions or producing and selling products according to consumer requirements. However, the concept and scope of customized cosmetics are not clear, such as various types of products can be made according to technological development and customer needs, and the lack of relevant legal basis has raised the problem that it may act as an obstacle to the growth of the industry. Accordingly, the Ministry of Food and Drug Safety promoted the customized cosmetic pilot project in March 2016 to revitalize the customized cosmetic industry. Since then, it is trying to prepare an institutionalized plan for the stable settlement of customized cosmetic products and to establish an effective management system. . In March 2019, the Ministry of Food and Drug Safety announced the implementation of the customized cosmetic system by March 2020 while enforcing some amendments to the Cosmetics Act. We are promoting the establishment of measures such as preparation of qualification tests, sales floor facilities and safety management standards.

Overseas, customized cosmetics are managed according to the general cosmetics laws, and there are currently no countries that have a management system specialized for customized cosmetics, but in Europe and the United States, there are regulations on sellers or similar guidelines. Although there are no detailed regulations on the sale of customized cosmetics in Europe and the United States, the reason that the sale of customized cosmetics in the form of on-site mixing is possible is not a pre-approval system in the relevant country, and the company is solely responsible for the exclusion of prohibited ingredients. It seems that this is because the ingredients can be mixed and used under the

All member countries belonging to the European Union are subject to the European Union Cosmetic Regulation (EC Cosmetic Regulation 1223/2009), and all cosmetics that enter and distribute in the European Union market must comply with the regulations. Although the subject of regulations under the Cosmetics Regulations of the European Union and the Korean Cosmetics Act are different, cosmetics for basic, color, and fragrance, which are generally sold as customized cosmetics, are classified and managed as cosmetics in both Europe and Korea. For cosmetics sold in Europe, a person in charge must be set for each cosmetic product for a clear regulation of responsible material, and it is stipulated that the person in charge is ultimately responsible.

Cosmetics imported and distributed in the U.S. are governed by the Federal Food, Drug, and Cosmetic Act (FD&C Act), The Fair Packaging and Labeling Act (FP&L Act), and the federal jurisdiction of the U.S. FDA. It is subject to the regulations, and customized cosmetics are also subject to applicable laws just like general cosmetics. Except for sunscreens managed as pharmaceuticals in the United States, cosmetics for basic, color, and fragrance products sold as customized cosmetics are classified and managed as cosmetics in the same way as in Korea. According to the U.S. Cosmetics Regulations, companies or individuals that manufacture and distribute cosmetics are responsible for labeling and ensuring safety of cosmetics, and in case of problems with cosmetics, the company has to bear responsibility according to the Product Liability Act. Cosmetics raw materials are basically managed on the Negative List (free use except for prohibited and restricted substances) except for colorants, and the Positive List (used only for approved items) is applied for colorants.

According to Article 13 of the Law for Ensuring Quality, Efficacy, and Safety of Drugs and Medical Devices in Japan, only those who have obtained the manufacturer's license can manufacture cosmetics under GMP compliance. It is difficult to directly manufacture cosmetics in the field because the shipment decision is determined by testing and inspection after container filling. Since cosmetics in Japan have to attach only one product name to a single manufacturing method, there is a problem that the manufacture of cosmetics tailored to the consumer must be shipped with different sales names, so customized cosmetics will be virtually impossible.

The present invention intends to propose a mobile skin imaging device and image analysis system developed including the following technologies.

- Technology that utilizes general visible and ultraviolet light when shooting skin images

- Technology to mount a polarizing filter to observe skin characteristics

- Communication technology to transmit/receive data between diagnostic devices and mobile environments

- Diagnostic device system technology

- Technology to control diagnostic devices through applications in a mobile environment

- Technology for matching and diagnosing skin characteristics and symptoms according to the standard data quantitative value of diagnostic items

As the main technical characteristics according to the H/W characteristics, in terms of hardware, it is largely composed of a lens, lighting, camera module, image sensor, image processor, communication module (Wi-Fi), and battery (charging). is controlled through the APP. More specifically, it utilizes multiple LED light source (normal light/polarized light/UV light) filters and a sensor function module for moisture content measurement, and is a Wi-Fi communication method (interlocking with mobile devices such as smartphones and tablets).

The reason for using multiple light sources is that when only a simple light source (general visible light) is used for skin imaging, there is a limit to the differentiation of skin characteristics, so the light source is subdivided for product differentiation and refinement of results. (Example: A polarizing filter is mounted to observe the pigment characteristics of the dermis of the epidermis of the skin, and ultraviolet (UV) light is mounted to obtain a reflection image such as excess sebum)

As a major technical feature according to the characteristics of the S/W algorithm, in terms of software, image processing techniques are applied to images taken in an analysis method, and definitions for analysis are made based on skin troubles and patterns expressed in images. That is, by recognizing patterns such as color/shape/thickness/direction of an object, it is possible to quantify the ratio of a specific pattern (feature) to the observation range. On the other hand, moisture can be detected by using an electrical resistance (impedance) measuring sensor of the skin epidermis.

Regarding the mobile device, there is an unmanned facial and hand skin analysis device for beauty that can measure and diagnose the skin condition of the face and hands by taking a picture of the face or hands alone using a kiosk. More specifically, when waiting in a place where many people gather, the user simply diagnoses the skin condition during the waiting time, and the measured diagnosis result data is transmitted directly from the unmanned facial and back skin analysis device to the user's mobile device. As the user (U) proceeds with a new skin diagnosis, the unmanned facial and back skin analysis device provides a new skin diagnosis result of the user (U), a previously stored skin diagnosis result of the user (U), and a new skin diagnosis of the user (U). The comparative information of the result may be included in the skin diagnosis result and transmitted to the user's mobile terminal. At this time, the communication module can transmit the skin diagnosis result to at least one of the user's mobile device, e-mail, SMS, and SNS through the wired or wireless network, and the communication method can use Ethernet, WIFI, LTE, 3g, 4g, 5g. have.

Meanwhile, in the present invention, the mobile device corresponding to the analyzer receives the image data of the skin photographed by the diagnostic device H/W in real time, and then analyzes the image data received through the dedicated application (APP) installed in the mobile device. / By operating the analysis algorithm, information indicating skin characteristics (troubles) is collected, patterns expressed in troubled skin images are recognized, quantified and quantified, divided into categories, and classified into various skin types and labeled. That is, in the present invention, input, processing, analysis, classification, and output of skin image data are processed using a predetermined APP in a mobile terminal owned by the user, and the final result after shooting and analysis in a centrally fixed device However, there is a difference in the technical configuration from transmitting only to the user's mobile device, and it will be said that the method as in the present invention has a higher degree of freedom from the user's point of view and a large scalability that can be applied in terms of S/W.

In addition, it includes a wireless communication module for transmitting the shape information of the skin disease target surface to a smart device (mobile device) through wireless communication. Therefore, the image shape information of the skin disease target surface stored in the memory module is transmitted to the smart device through wireless communication. It is transmitted to the device, and the user discloses a configuration in which the image information of the target surface of the skin currently being photographed is checked in real time through the screen of the smart device. In addition, by composing a three-dimensional stereoscopic image and providing it to a remote cloud server, the skin disease target surface diagnosis information is comprehensively generated in the cloud server. Accordingly, users can acquire the skin diagnosis information by accessing the cloud server anytime, anywhere through a smart device having the skin disease surface diagnosis information. To this end, a diagnosis app (APP) is installed on the smart device and connected to the cloud server through this, so that a more precise skin condition can be diagnosed and skin diagnosis errors can be prevented.

The cloud server includes a machine learning skin disease learning unit to learn and set the number of images related to skin diseases (for example, skin cancer) using machine learning techniques, so that the difference in absorption rate and the optimal absorption rate in the collected skin disease images are learned. Thus, the accuracy of the diagnosis of skin diseases can be improved.

This cloud server acquires thousands to tens of thousands of skin disease image information through the machine learning skin disease learning unit, analyzes the absorption rate for each wavelength band of skin diseases, especially skin cancer, and sets the absorption rate threshold difference through this. By providing it as a diagnostic app for smart devices, it is possible to diagnose whether or not skin cancer is present with the difference in absorption rate determined by the smart device. In addition, according to an additional modification, the cloud server may further include an expert diagnosis result obtaining unit for requesting a skin cancer visual diagnosis result by providing skin cancer image information obtained by a doctor who is an expert group through the terminal. That is, the diagnosis accuracy of skin cancer can be further increased by obtaining the diagnosis result of skin cancer through the visual examination information of doctors. In the present invention, in developing a dedicated diagnostic application (APP) for a mobile device that performs a skin analysis algorithm, the above cloud server and AI module (machine learning algorithm) can be applied. Various patterns expressed in skin images including skin characteristics (troubles) can be machine-learned, converted into a DB in a cloud server, and an analysis algorithm can be automated using this. The accuracy of skin analysis can be further improved by adding a DB that reflects the comments of experts such as dermatologists.

As a method for categorizing images by pattern analysis, when skin image information of a member is received from a skin analyzer, the skin condition is classified into a predetermined pattern through a classification algorithm, and after relatively comparing the skin condition with that of other people, the skin It can be configured as a server that predicts the skin condition in advance through time series analysis of , and recommends cosmetic information corresponding thereto. The camera unit magnifies the skin condition up to 500 times and shoots in a contact or non-contact manner to detect pores, hair follicles, demodex, skin pores, hair, dead skin cells, capillaries, hair, spots, freckles, blemishes, skin types, wrinkles, etc. Long-distance shooting function that takes detailed shots and non-contact full face and full body shots, and close-up shots for each part of the face, enables contact-type short-distance shooting to compare before and after. The photographed image information is received by the skin analysis unit, and the skin analysis unit analyzes the member's skin image information to calculate the indices for moisture, oil, elasticity, skin texture, wrinkles, pigmentation, pores, sebum, trouble, and sensitivity, respectively. divided into a certain pattern. At this time, the skin condition of the member is relatively compared with the skin condition statistical information of other members using a fuzzy algorithm, and the skin condition of the member, which changes over time, is relatively compared. can be predicted As described above, in the process of calculating each index by analyzing skin image information and dividing it into a certain pattern, (a) the method of indexing the moisture measurement result of the skin is to set the level of moisture that the consumer's skin can hold from 1 to 100, and From this, when the actually measured moisture is in the range of 1 to 30, "excessive moisture", when it is in the range of 31 to 50, "slightly lacking moisture", when it is in the range of 51 to 75, "slightly lacking moisture", 76 to When it is in the range of 100, it can be judged as "moisture titration", and in this similar way, other types of skin conditions such as oil and elasticity, skin texture, pigmentation, pores, troubles, and sensitivity can also be indexed. have. On the other hand, (b) it is desirable to measure the skin texture in detail, taking into account the density of the skin texture, the skin preparation index (roughness-vertical density), and the evenness of the skin pores, (c) the wrinkles around the eyes, (d) Pigmentation can be measured in more detail taking into account the depth of wrinkles around the mouth and the depth of forehead wrinkles, and (d) Pigmentation can be measured in detail considering the distribution of surface pigments, distribution of inner pigments, and the degree of progression of inner pigments. (e) Pores can be measured in consideration of pore size and pore area, and (f) Sebum can be measured in consideration of accumulated sebum and general sebum. And (g) trouble can check whiteheads, purulent acne, blackheads, acne, pimples, etc. through the naked eye of the camera depending on the type of trouble, and (h) sensitivity is itching, keratin generation, fever redness, visible blood lines, For flushing, etc., input through the questionnaire, and the time of sensitivity occurrence can be indexed as described above by considering whether it occurs at a specific time, periodically, or routinely.

In performing the skin analysis algorithm in the present invention, oil, pores, wrinkles, pigmentation, sensitivity, sebum, and moisture (more than 7 types) are dealt with as exemplary diagnostic items, and each measurement method includes oil occupancy area, number of pores, It can be defined by measuring the wrinkle area, the pigmentation area, the sensitivity area, the number of sebum, and the water impedance sensor value. In developing and confirming an analysis algorithm that analyzes image information and indexes and patterns it into various items, a formalized method may be applied.

As for the business model (BM) related to customized cosmetics, the manufacturing of cosmetics according to the prior art was mostly factory-style mass production of cosmetics with the same ingredients and the same capacity in the same packaging container. In particular, cosmetics that are suitable for sensitive skin and have no side effects had to be tailored individually. However, due to the recent spread and development of 3D printers, 3D printing technology has been applied to construction, automobile parts, and small-scale production of compounds having different chemical composition ratios, so that it can be used in personal cosmetics manufacturing equipment.

Cosmetic manufacturing apparatus according to an embodiment relates to a technology for manufacturing cosmetics suitable for individual skin characteristics using a printer based on individual skin analysis, and specifically (a) a plurality of cosmetic raw materials equipped with a pressure regulator (b) a nozzle that receives cosmetic raw materials from the container through a tube and sprays it into the cosmetic container, (c) a valve device that controls the amount of cosmetic raw materials discharged through the nozzle, and (d) a wood lamp Receives a control command to control the pressure regulator, nozzle and valve device from the user terminal to adjust the blending ratio of cosmetic ingredients based on skin condition information analyzed as an image for skin analysis taken using a built-in luminescent mirror. It includes a control board, (e) characterized in that it further comprises a sensor for detecting the amount of the cosmetic raw material delivered through the tube.

On the other hand, the cosmetic manufacturing method based on skin analysis using a luminescent mirror according to the present invention is (a') skin analysis using a luminescent mirror that provides a light source through each or a combination of ultraviolet rays, polarized light and white light. (b') performing skin analysis through a user terminal based on the obtained image, and (c') a cosmetic device for manufacturing cosmetics based on skin condition information according to the analysis result. It characterized in that it comprises the step of manufacturing cosmetics using. In the present invention, a mobile device receives image data of skin photographed by a diagnostic device in real time, and by analyzing the received image data through an analysis algorithm of a dedicated application, image information representing skin characteristics is recognized as a pattern and digitized by category. It is about an analysis method that divides and classifies into various skin types, and by using the analysis results processed in this way, we intend to manufacture personalized cosmetics and sell them to general consumers.

According to another embodiment of the present invention, in terms of H/W of the diagnostic device, it is not only subdividing the light source (normal light/polarized light/UV light) to overcome the limitation in skin characteristic pattern differentiation, but also a RAW image that makes it easy to detect skin troubles. Various photographing methods can be applied for acquisition. (Example: parallax shooting, low-light shooting, zoom shooting, etc.)

According to another embodiment of the present invention, various auxiliary sensors capable of compensating for information errors of the original photographed image data may be utilized. (Example: temperature sensor, heart rate sensor, other contact/non-contact biosensor, etc.)

According to another embodiment of the present invention, in terms of the analyzer S/W, an algorithm including an equation may be devised and an optimal value (or an optimal formula) of the skin diagnosis criterion may be derived through sufficient real-case tests.

According to another embodiment of the present invention, in terms of grafting with AI, a method of grafting big data (including expert DB) and machine learning algorithms and deep learning technology may be added.

According to another embodiment of the present invention, in terms of the BM model, as a step linked to business after skin analysis and trouble diagnosis, not only customized cosmetics manufacturing but also linkage with medical facilities, beauty product shopping mall purchase, health care (disease prevention and Management), health food (nutritional supplements), skin care product evaluation/certification, personal broadcasting beauty channel (YouTube), daily self-diagnosis and makeup education, etc. are available.

According to another embodiment of the present invention, real-time feedback and mutual control scenarios between a diagnostic machine (photographing module) and an analyzer (image processing module) for a high-quality, high-performance diagnostic process in terms of a system in which H/W and S/W are combined Through this method, a more accurate diagnosis process can be built.

Claims (3)

In the customized skin diagnosis and image analysis system,
Using a plurality of light sources, the image photographed by the mobile skin imaging diagnostic device is transmitted to the mobile terminal, and the image pattern as a result of image processing the optically observed image in the diagnostic device in conjunction with the mobile terminal and the pre-installed app is displayed on the skin. A skin diagnosis and image analysis system capable of producing personalized cosmetics according to the typology by classification and matching according to a predetermined characteristic, and then according to the typed analysis result.
According to claim 1,
The plurality of light sources are general visible light, polarized light, and ultraviolet light sources, which are integrated into the mobile dermatography diagnostic machine and analyze the state of sebum, wrinkles, pores, black spots, erythema, blackheads or keratin of the enlarged skin image based on this ,
The mobile skin imaging diagnostic device simultaneously includes a moisture sensor to provide a high-resolution skin diagnostic device function that measures the level of moisture by measuring the impedance of the skin epidermis, and controls it through data communication and application for interworking with the mobile environment with the mobile terminal is possible,
The mobile terminal and the pre-installed app apply an image processing algorithm to the captured image to establish a definition for analysis based on the pattern expressed in the image, and then recognize and observe the pattern of color, shape, thickness or direction of the object. A skin diagnosis and image analysis system that quantifies the ratio of a predetermined pattern feature to a specified range, and classifies, matches, and types the characteristics obtained by analyzing the skin image image as a pattern.
According to claim 1,
The mobile skin imaging diagnostic device consists of a lens, a lighting, a camera module, an image sensor, an image processor, a communication module, and a battery in hardware,
Specifically, in order to subdivide the light source, various types of LED light source filters of general light, polarized light and UV light and a sensor function module for measuring moisture content are used, and it is a method of interworking with a mobile terminal through Wi-Fi communication,
The mobile terminal sets definitions for analysis based on skin troubles and patterns expressed in images when applying image processing techniques to images captured by software, and image data of skin photographed by the mobile dermatography diagnostic device is received in real time, and by applying an analysis and interpretation algorithm to the received image data through a pre-installed dedicated application, information indicating skin characteristics such as skin troubles is collected, and patterns expressed in trouble skin images are recognized and processed. It is characterized by quantifying and quantifying, dividing by category, and classifying and labeling multiple skin types
The cloud server includes a database set by learning the number of images related to skin diseases using machine learning techniques, so that the cloud server can be accessed through the mobile terminal to obtain skin diagnosis and image analysis. system.

Images