VIRTUAL MAKEUP SIMULATION SYSTEM BASED ON RELIABILITY-WEIGHTED FUSION OF MULTIPLE VISUAL DATA AND FEEDBACK-BASED VISUAL CONSISTENCY CHECK

Abstract

MAKEUP SIMULATION SYSTEM The invention relates to the Multiple Visual Data Processing and Feedback Virtual Makeup Simulation System (1) and its method, which is based on processing multiple visual data obtained from the same user under different viewing and lighting conditions. In the visual data obtained by the visual data acquisition unit (2), face and skin regions are identified, the corresponding regions are matched by the inter-regional image matching unit (8), and regional surface features are extracted by the surface feature extraction unit (13). The variability of these features between different visual data is determined by the inter-image feature variability determination unit (17), the regional feature reliability value (19) is created by the regional feature reliability determination unit (18), and the surface features are weighted according to the reliability values ​​by the reliability weighted feature combination unit (20).The User-Specific Surface Feature Data Structure (22) is created by the Stable Surface Feature Creation Unit (21); the Regional Image Synthesis Parameters (29) are determined by evaluating this data structure together with the visual context features. Using the Target Lighting Data (31) created by the Target Lighting Modeling Unit (30), the Synthesized Virtual Makeup Image Synthesis Unit (32) creates the Synthesized Virtual Makeup Image (33). The Regional Visual Consistency Control Unit (34) determines the Regional Visual Consistency Value (35), which represents the region of the discrepancy, the type of image feature, and the amount of discrepancy in the synthesized image; the Discrepancy Classification and Parameter Matching Unit (36) matches the discrepancy with the relevant synthesis parameter.Based on the correction data generated by the inconsistency-related parameter correction data generation unit (37), the Feedback image synthesis parameter update unit (38) updates at least one of the relevant Regional image synthesis parameters (29), and the resynthesis process is controlled by the Consistency condition evaluation and resynthesis control unit (39) based on the consistency condition.

Description

1 TARIFF RELIABILITY-WEIGHTED FUSION OF MULTIPLE VISUAL DATA. AND BASED ON FEEDBACK-BASED VISUAL CONSISTENCY CHECK VIRTUAL MAKEUP SIMULATION SYSTEM Technological Field: 5 This invention involves computer-assisted image processing, computer vision, and multi-image analysis. Image feature extraction, image data merging, digital imaging It relates to the field of synthesis and virtual appearance simulation technologies, and especially different Face and Extraction of regional surface features relating to skin regions, 10 of these features Determining their reliability and combining them based on reliability values. computer-generated surface feature data structure for user-specific purposes. It relates to an applied system. More specifically, the invention relates to corresponding faces and skin within different visual data. Matching of regions, regional characteristics resulting from imaging conditions 15 Determining the variations in color, tone, and brightness extracted from the regions in question, Usability and reliability information of contrast and surface appearance properties creation, reliability-weighted feature fusion, and the resulting User-specific surface feature data structure along with visual context features. By processing, the image synthesis parameters specific to facial regions are determined. 20 This includes related image processing and data fusion technologies. The invention also allows for virtual image synthesis using specified regional image synthesis parameters. Synthesizing the makeup look depending on the target lighting conditions, Checking the regional visual consistency of the synthesized image, the resulting Determining the location, type and quantity of discrepancies, the identified 25 matching discrepancies with related image synthesis parameters and wording The virtual makeover is updated based on the discrepancy data of the parameters in question. feedback-based and closed-loop imaging for image resynthesis It is related to processing technologies. State of the Art: 30 Today, computer-assisted image processing, computer vision, augmented reality, due to advancements in machine learning and digital image synthesis technologies Analyzing users' facial and appearance features for cosmetic applications. 2 the creation of related proposals and the presentation of the created views to the user in a digital environment. Various systems and methods are being developed for displaying the image on a screen. These systems generally use data from a camera or similar imaging unit. The user's image is being processed, and the face and related areas within the image are being identified. Various visual features are extracted from these regions, and the resulting data is used for makeup application. It is used for selection, makeup suggestions, or creating a virtual look. In some applications under the known state of the art, the user's facial image by taking the image and analyzing the facial features in the image in question Choose a suitable option from predefined makeup stencils or cosmetic choices. The issue is choosing an option that is visible. The chosen makeup look is 10... application to the relevant areas of the face, color information regarding cosmetic products usage, identification of application areas on the face and the creation The view is known to be displayed to the user either progressively or in real time. It is among the applications. Some systems consider not only the user's facial features but also their clothing, body, or 15 appearance characteristics, color information, environmental conditions, user preferences, and similar additional factors. Data is also taken into consideration. Using this data, the user is provided with... Appropriate makeup or clothing options can be identified and obtained. The available options can be presented to the user in a digital environment. Therefore, face-to-face Features derived from the image, along with clothing, appearance, or environmental data, 20 This evaluation is among the known approaches in the technical field. Utilizing learning-based data processing approaches in the known state of the art Systems also exist that extract features from facial images. The trained algorithm uses data about the user, their appearance, or other surroundings. The data can be transferred to models and, as a result of the data processing performed, makeup or 25 Appearance suggestions can be generated. Therefore, facial features can be determined from the image. extraction, combined evaluation of different user and contextual data, and word the subject matter is a makeover or transformation of data using algorithmic or learning-based models. Linking the appearance to the result is also a known practice in the technical field. In addition, there are 30 virtual representations that represent the user's physical or visual characteristics. the creation of models and the selection of clothing, hair, makeup or similar appearance elements There are systems for visualizing this subject using models. These types of systems... a digital representation of user-related image or measurement data in buildings 3 It can be created and different view options are offered to the user on that representation. This allows the user to be presented with their real image or a virtual representation. Application of different visual elements and evaluation of results in a digital environment. It is possible. The approaches in question under the known state of the art include: facial image acquisition, facial 5 Identifying features, analyzing the user's different appearance characteristics, Creating makeup recommendations involves evaluating clothing or environmental data. the acquisition, the use of learning-based models and the resulting view It provides various solutions in terms of being presented virtually to the user. together display of visual data used in carrying out these operations 10 the surface to be used in creating the virtual view, being affected by the conditions This creates a technical problem in terms of the stability of its properties. Images belonging to the same user may be displayed at different times or under different viewing conditions. When obtained, factors such as light intensity, direction of light, color temperature, camera angle, and exposure are considered. Factors such as shadowing, reflection, and image quality can vary. The 15 in question... Color, tone, brightness, contrast of the same face or skin area due to changes And surface appearance characteristics can show different values ​​in different images. Therefore, a feature removed from the image is in any case only the physical attribute belonging to the user. or does not represent the visual surface characteristic, the conditions under which the image was obtained It can also include the variability it creates. 20 This is particularly important in terms of virtual image synthesis. Imaging a temporary change in color, brightness, or shade resulting from the condition is the responsibility of the user If this variability is considered as a stable surface property This can be transferred to the determination of image synthesis parameters and the generated This can cause the virtual appearance to deviate from the actual user characteristics. 25 In structures where a single image is used, excessive or insufficient lighting, regional Effects such as shadowing, glare, reflection, exposure variation, or chromatic aberration are specific This can reduce the usability of features extracted from facial regions. Using multiple images together also solves the problem itself. It does not eliminate the features extracted from different images. Displaying 30 without determining the level of impact from the conditions, directly or with equivalent weight If combined, data extracted from a region with low availability 4 The surface representation is created based on data obtained under more favorable viewing conditions. They can be equally effective. In addition, different facial regions within the same image are subject to equivalent imaging conditions. It may not have this. While one area of ​​the face is properly illuminated, another area may not be. The area may be in the shade, affected by reflection, or excessively bright. It can be exposed. Therefore, only an overall quality of the image is possible. the evaluation of features extracted from different facial regions It may not represent its usability with sufficient accuracy. Another technical problem encountered when using multiple visual data together is: 10 data corresponding to the same anatomical or surface region within different images It is associated with the user's head position, viewing angle, facial expression, or When the scale of the image changes, a specific facial region in the image is extracted. direct comparison of the feature with the corresponding region in the other image This can make it difficult. Therefore, the corresponding face and skin areas in different images matching and feature comparisons are performed through these matched regions 15 Its implementation is of technical importance. After identifying the corresponding regions in different images, the aforementioned the variability of features extracted from regions between images Evaluation also presents a separate technical need. Throughout the different images Regional features showing similar values ​​provide a more stable surface for the user. 20 While it can represent its characteristic, it is significant only in certain images. changes in lighting, shadowing, reflection, exposure, or other imaging This may be due to the conditions. Therefore, the difference between images variability, along with viewing conditions and regional availability information. It needs to be evaluated. 25 In this context, all features extracted from different images are presented in an equivalent manner. instead of using a system that represents the availability of each regional feature Establishing reliability information is of technical importance. This reliability... the display conditions of the information, regional image availability and inter-image 30 Its use in combining images is heavily influenced by imaging conditions. This allows for limiting the impact of the data on the generated surface representation. will be able to give. Another need in the technical field is to obtain images from different sources with varying reliability. Instead of simply selecting or eliminating regional characteristics with varying levels, reliability information as a parameter of the feature combining operation. This involves using features that have higher reliability. Increasing the contribution of surface representation, features with lower reliability 5 by reducing its contribution, the user is given more consideration for the viewing conditions. It will be possible to create a stable surface feature data structure. Another technical problem encountered in creating a virtual appearance is the face. This refers to the fact that different regions of the forehead have different surface and visual characteristics. Color distribution and tone of cheeks, eye area, nose area, lips, and other facial areas, 10 in terms of brightness, contrast, surface appearance, and visual response to lighting This can vary. Therefore, the entire face may not have a common or pre-existing appearance. Application of defined, invariant image synthesis parameters, regional the features are not adequately represented and the virtual appearance created does not look natural. This can lead to the occurrence of non-existent regional transitions. 15 On the other hand, the surface characteristics of the user and the color distribution in the clothing areas and Contextual data, such as visual characteristics relating to the viewing environment, is only one The use of makeup to determine or suggest makeup options, This limits the technical use of the data in the image synthesis process. Different The relationship between data groups is directly related to face region-specific image synthesis 20 its use in determining parameters, visual context is only a suggested input. from being a tool to becoming a technical parameter of virtual image production It creates a need in terms of maintenance. Lighting variability is also a technical issue encountered in virtual reality rendering systems. This is one of the problems. The color, tone, and intensity created under specific lighting conditions... 25 and the transfer of brightness values ​​to different lighting conditions without changing them, This causes the generated virtual view to become inconsistent with the target viewing conditions. This is possible due to factors such as light intensity, the direction of light incidence, and the color characteristics of the light. Because changes can produce different visual effects in different areas of the face, Determination of regional image synthesis parameters under target lighting conditions 30 and needs to be included in the process of creating the virtual appearance. After the virtual view is created, the resulting image is technically... Monitoring is also a separate problem area. The facial regions are the result of the synthesis process. 6 Sudden changes in color, tone, or intensity between colors, incongruous transitions between adjacent areas, Brightness values ​​that are incompatible with the target lighting conditions or user-specific regional image characteristics inconsistent with surface features are produced. It is possible. The generated image is only presented to the user and synthesis is performed. In structures where the result is not re-analyzed, 5 errors occur during the synthesis phase. automatic identification of inconsistencies and the synthesis process in question Modification based on incompatibilities may be limited. Therefore, the synthesized virtual image is re-analyzed in different facial regions. the processing of regional image characteristics, user-specific surface features and target Comparison with lighting conditions and detection of any discrepancy that occurs. 10 Instead of determining it, there is a technical need to determine it in terms of location, type and quantity. It is located there. The assessment in question concerns which region of the image and which This allows for determining when corrections are needed in terms of image quality. will be able to give. In addition, the identified non-conformity is only a result of a control. instead of creating, the image synthesis parameter with which the mismatch is associated identification and discrepancy information in changing the parameter in question It is necessary to use it. In this way, regional color, tone, brightness, and intensity, A mismatch related to one of the transparency, distribution, or transition properties, Converting the correction data corresponding to the synthesis parameter and the image 20 It may be possible to resynthesize it with updated parameters. The result obtained after the image is resynthesized is repeated. evaluation and checking whether a specified consistency condition is met. This is also important for the stability of the image synthesis process. Inconsistency detection if the consistency condition is not met, parameter 25 The ability to repeat correction and resynthesis processes, image synthesis. instead of a one-way data processing chain, the process involves the previous processing stages of the result data. it allows it to be carried out within a closed-loop structure that it has modified. It will be able to provide. In conclusion, in the current state of the technique, facial image analysis involves 30 faces. Identifying features, creating makeup suggestions, clothing and other visuals. evaluation of contextual information, utilization of learning-based models, virtual makeup application and the resulting look under different viewing conditions 7 Although various technical solutions exist for delivering it to the user; different Corresponding face and skin in multiple visual data obtained under imaging conditions. matching regions, extracting features from matched regions, and cross-image mapping. Determining their reliability based on variability and regional availability, These features are weighted and combined based on their reliability values, resulting in 5 User-specific and more stable surface feature data depending on viewing conditions. the creation of the structure, based on the data structure and visual context features in question Generating region-specific image synthesis parameters, targeting the virtual view Synthesis depending on lighting conditions, regional variation in the synthesized appearance Determining the location, type, and quantity of discrepancies, as determined in the 10 Parameter correction by matching mismatches with related synthesis parameters. conversion of data and re-processing with updated parameters the synthesis process is repeated depending on a specified consistency condition a need for an integrated and feedback-based image processing structure that enables this It is located. 15 The purpose of the invention: The purpose of the invention is to obtain multiple views under different viewing and lighting conditions. visual data is treated only as independent images or words The subject matter differs from structures that use features extracted from images in an equivalent way. Specifically, identifying corresponding facial and skin regions within different visual data, the word 20 the effect of imaging conditions on surface features extracted from the regions in question by evaluating levels, regional characteristic reliability values ​​are created and different The features obtained from the images are weighted according to the aforementioned reliability values. by combining them in a way that is user-specific and responsive to changes in viewing conditions a computer-implemented virtual 25 that creates a more stable surface feature data structure The goal is to develop a makeup simulation system. One aim of the invention is to identify facial and skin regions within different images. matching them with each other and images relating to the same or corresponding regions. The aim is to enable comparison of features between images. In this context, different Color, tone, brightness, contrast, and surface 30 extracted from corresponding regions in the images Determining the variability of visual characteristics between images and the aforementioned the use of variability in assessing the reliability of regional characteristics that is intended. 8 Another purpose of the invention is to show face and skin areas in different images only in general. not according to image quality, but separately according to regional viewing conditions. The aim is to enable its evaluation. In this regard, lighting characteristics, exposure, such as shadowing, reflection, color stability and regional image usability. 5 different facial regions in different images by evaluating the factors. regional usability of extracted features in the virtual reality creation process The aim is to determine this. Another aim of the invention is to improve imaging conditions, regional image availability, and from different face and skin regions, taking into account feature variability between images 10. Creating regional feature reliability values ​​for the extracted features The aim is to ensure that the features obtained from different images belonging to the same user are... to prevent them from being considered equivalent simply because they have been obtained and the degree to which each feature is affected by display conditions in the virtual view The aim is to transfer this to the creation process. Another aim of the invention is to identify regional features derived from different images. weighting the reliability values ​​of the regional characteristics in the combination The aim is to use it as information. In this regard, regional options with higher reliability values ​​are preferred. Increasing the impact of features on the combined surface representation, reliability value reliability is reduced by limiting the impact of lower regional characteristics. The aim is to perform a weighted feature combining operation. 20 Another aim of the invention is to create different features through a reliability-weighted feature combination process. intermittent transitions between images resulting from viewing conditions. Represents more stable surface characteristics for the user compared to variations. The aim is to highlight the defining characteristics of lighting and shading. This allows for the clarification of these features. Temporary image resulting from reflection, exposure and similar imaging conditions 25 the effect of their variability on more stable surface properties for the user reduction and creation of stable surface properties for different surface regions that is intended. Another purpose of the invention is to combine regional characteristics with a focus on reliability. To create a user-specific surface feature data structure and 30 virtual makeup images in its synthesis, instead of the instantaneous visual characteristics of a single image, the following is used: The goal is to use a user-specific surface feature data structure as technical input. In this way... color, tone, brightness, contrast, and surface appearance of different areas of the face 9 features that are regionally represented and obtained from different viewing conditions providing a surface representation in which the collected data is stored within a common structure that is intended. Another aim of the invention is to create a user-specific surface feature data structure for clothing. color distributions in the regions and visual characteristics related to the viewing environment 5 The goal is to create a unified context-attribute data structure by working together. Within this scope... Clothing and environmental factors only suggest a makeup option to the user. Instead of being considered as supplementary information, user-specific surfaces along with their features, become the technical inputs of the virtual image synthesis process. The aim is to bring it about. 10 Another objective of the invention is to provide a user-specific surface feature data structure with integrated context. By evaluating the feature data structure together, characteristics specific to different regions of the face are identified. The aim is to enable the determination of image synthesis parameters. This will allow for the facial recognition of the image. Instead of applying common and unchanging image synthesis values ​​to all of them, Color, tone, brightness, density, transparency, regional distribution and 15 for different facial areas regional image synthesis parameters that identify at least one of the transition characteristics The aim is to create it. Another aim of the invention is to create images that are specific to different regions of the face. synthesis parameters that preserve the visual relationship between neighboring regions The aim is to ensure that it can be used in a way that allows for this. In this regard, 20 different facial regions... The applied virtual makeup components change color, tone, and intensity suddenly along regional boundaries. or limiting the occurrence of transparency changes and different regions visual transitions between them created in a way that is consistent with regional surface features that is intended. Another objective of the invention is to create a virtual makeup look using only 25% of the source visual data. Instead of creating lighting based on existing lighting conditions, targeted lighting is created. The goal is to make these conditions one of the technical inputs of the image synthesis process. within the scope, it represents the light intensity, light direction, and color characteristics related to target lighting. the creation of data and the determination of this data for different facial regions Virtual makeup 30 by evaluating it together with image synthesis parameters The aim is to synthesize the image depending on the target lighting conditions. Another objective of the invention is to create a user-specific surface feature data structure, regional image. Virtual lighting is created using synthesis parameters and target illumination data. a one-way process resulting in the makeup look being presented solely to the user instead of an image processing process, the synthesis result is technically re-analyzed. The goal is to create an image processing structure. In this context, the synthesized virtual makeup... re-analysis of different facial regions of the image and regional color, tone, The aim is to determine brightness, intensity, distribution, and transition properties. 5 Another purpose of the invention is to identify regional features from the synthesized virtual makeup image. the features include a user-specific surface feature data structure, target lighting data, and related information. regional visual by comparing with regional image synthesis conditions The goal is to perform a consistency check. This way, the results obtained from the synthesis process... 10 possible regional color, tone, brightness, intensity or transition inconsistencies It is intended to be determined automatically. Another purpose of the invention is to identify a result of regional visual consistency audit. Instead of viewing incompatibility solely as knowledge of existence or non-existence, the word the facial region where the discrepancy occurred, and the associated visual feature and a regional visual discrepancy value of 15 representing the amount of discrepancy. The aim is to enable its creation. In this way, it can be determined in which region of the virtual image, which quantitatively, in terms of image quality and the magnitude of correction required. The aim is to determine this. Another purpose of the invention is to classify the type of identified regional visual discrepancy. and 20 The goal is to correlate this with the regional image synthesis parameter that needs to be modified. Thus, for example, color, tone, brightness, intensity, transparency, regional distribution, or transition can be affected. a discrepancy determined in terms of its characteristic related to image synthesis matching with the parameter and recalculating the mismatch instead of a general recalculation. The aim is to determine the parameter. 25 Another purpose of the invention is to address the location, type, and amount of regional visual discrepancy. the subject depends on the image synthesis parameter matched with the mismatch The aim is to enable the generation of parameter correction data due to discrepancies. correction aimed at reducing the discrepancy identified within the scope of the relevant facial area and its definition in terms of the relevant image synthesis parameter and the aforementioned 30 The aim is to feed the correction data back into the image synthesis process. Another purpose of the invention is to provide parameter correction data resulting from discrepancies to the relevant regional context. by feeding feedback to the image synthesis parameters, at least of those parameters 11 updating one of them and the virtual makeup look with updated parameters The goal is to enable its resynthesis. In this way, the determined image is synthesized. Instead of simply reporting the technical incompatibility, the subsequent investigation of that incompatibility should be carried out. The aim is to directly modify the inputs of the image synthesis process. Another objective of the invention is to use updated image synthesis parameters to 5 regional visual consistency on the recreated virtual makeup image repeat the inspection and current regional visual discrepancies. The aim is to ensure the redefinition of their values. The current values ​​determined within this scope... whether the discrepancy values ​​satisfy a defined consistency condition assessment and non-compliance classification if the condition in question is not met, 10 parameter matching, parameter correction, parameter updating, and resynthesis The aim is to repeat the processes. Another purpose of the invention is to perform image synthesis and visual consistency checking processes. by moving away from independent processes, the results obtained as a result of synthesis 15 that changes the parameters to be used in the next synthesis cycle of the image data The goal is to create a closed-loop image processing structure. This way, the system only... not as an image processing system that creates a virtual makeup look, but as what it creates assessing the regional technical consistency of the image, identifying inconsistencies, and related issues. correlated with synthesis parameters and providing feedback on those parameters. It functions as a system that reconstructs the virtual image by updating it in this way. 20 that is intended. The main purpose of the invention is, ultimately, to obtain multiple images under different imaging conditions. Matching the corresponding facial and skin regions within the visual data, inter-image variability of surface features extracted from regions and regional Determining their reliability, reliability of features obtained from different images 25 User-specific and display-based solutions are created by combining values ​​in a weighted manner. Creating a more stable surface feature data structure against these conditions, The data structure is processed together with visual context data to create images specific to facial regions. Determining synthesis parameters, target lighting of the virtual makeup image. Synthesis depending on the conditions, regional visualization of the synthesized image 30 Quantitative verification of consistency, related images of identified inconsistencies Synthesis parameter matching, parameter correction due to mismatch the creation of data and updating the relevant parameters through feedback to create a virtual system. 12 the makeup look must be reapplied until the defined consistency condition is met. an integrated, feedback-driven, and computer-implemented system that allows for synthesis The goal is to create a virtual makeup simulation system. Explanation of the Figures Figure 1: Multiple visual data processing and feedback virtual makeup simulation system; 5 from visual data acquisition to user-specific surface feature data structure creation, processing of visual context, regional image synthesis determining the parameters, virtual makeup image depending on target lighting synthesizing and returning the synthesized image as a result of visual consistency control. the overall system architecture and basic 10, extending to its reconstruction in a feed-in manner. A schematic view showing the data relationship between units. Figure 2: Multiple visual data obtained under different viewing and lighting conditions. Identifying the facial and skin regions within, and their corresponding appearances in different images. matching of regions, viewing conditions, lighting characteristics, exposure and determining shadowing and regional image availability, matched 15 extracting surface features from regions, and comparing these features between images. determining their variability and regional reliability from different images combining the obtained features in a weighted manner based on their reliability values ​​and As a result of this merger, the user-specific surface feature data structure Schematic view showing its creation. 20 Figure 3: User-specific surface feature data structure with visual context features. processing together, determining the clothing area and clothing color distribution, environment evaluation of visual characteristics, composite contextual features from the data in question the creation of a data structure, combined with a user-specific surface feature data structure. 25 regional image synthesis parameters depending on the context feature data structure. determination, generation of target illumination data and regional image Virtual makeup is created using synthesis parameters and target lighting data. Schematic view showing the synthesis of the image. Figure 4: Regional visual consistency of the synthesized virtual makeup image. monitoring, determining the regional visual incompatibility value, the determined 30 classification of the type of discrepancy and related regional image synthesis matching with the parameter, depending on the location, type and amount of the mismatch. generating parameter correction data, depending on said correction data 13 the relevant regional image synthesis parameter in a feedback manner updating, re-evaluating the virtual makeup image with updated parameters synthesis and monitoring until regional visual consistency is ensured, closed-loop system that allows for the repetition of correction and resynthesis processes. Schematic view showing the image processing structure. 5 References: 1. Multiple visual data processing and feedback-based virtual makeup simulation system. 2. Visual data acquisition unit 3. Multiple visual datasets 4. Image processing unit 10 5. Data storage unit 6. Facial region identification unit 7. Skin region partitioning unit 8. Inter-region image matching unit 9. Imaging condition analysis unit 15 10. Lighting feature determination unit 11. Exposure and shadow determination unit 12. Regional image availability determination unit 13. Surface feature extraction unit 14. Regional color feature data 20 15. Regional tone and brightness feature data. 16. Regional contrast and surface appearance feature data. 17. Unit for determining feature variability between images. 18. Regional feature reliability determination unit 19. Regional feature reliability value 25 20. Reliability-weighted feature combination unit 21. Stable surface feature formation unit 22. User-specific surface feature data structure 23. Visual context analysis unit 24. Clothing zone designation unit 30 25. Clothing color distribution data 26. Environmental visual feature data 27. Unified context feature data structure 14 28. Regional image synthesis parameter determination unit 29. Regional image synthesis parameters 30. Target lighting modeling unit 31. Target lighting data 32. Virtual makeup image synthesis unit 5 33. Synthesized virtual makeup image 34th Regional Visual Consistency Inspection Unit 35. Regional visual incongruity value 36. Incompatibility classification and parameter matching unit 37. Unit 10 for generating parameter correction data due to incompatibility. 38. Feedback image synthesis parameter update unit 39. Consistency condition assessment and resynthesis control unit. 40. Display and User Interaction Unit Description of the Invention: The invention involves obtaining multiple images under different viewing and lighting conditions. regional processing of data, corresponding faces and skin tones within different visual data. Identifying and matching regions, surface areas extracted from those regions the variability and reliability of their characteristics depending on the imaging conditions Determining the reliability of regional surface features obtained from different visual data. Combining them based on their values, the said combination 20 resulting from user-specific changes and variations in viewing conditions Creating a surface feature data structure in which transient effects are limited, The surface feature data structure is evaluated together with the visual context features of the face. Determining region-specific image synthesis parameters, target illumination The inclusion of these conditions in the image synthesis process, the virtual makeup image in question 25 the synthesis of the subject data, in the synthesized virtual makeup image Identifying regional visual discrepancies, and comparing the identified discrepancies with the relevant images. Matching with synthesis parameters, parameter correction for mismatches converting the data and synthesizing the parameter correction data in question. By feeding feedback through the process, the virtual makeup image maintains a defined consistency. 30 Multiple visual data processing and resynthesis depending on the condition. It is related to the fed virtual makeup simulation system (1). Multiple visual data processing and feedback virtual makeup simulation system (1), one or Capturing images of multiple users, identifying facial features, and by displaying a virtual makeup look based on specified facial features It does not create a limited data processing structure. Multiple visual data processing and feedback. Within the virtual makeup simulation system (1), 5 can be obtained under different viewing conditions. The corresponding facial and skin regions in the obtained visual data are related to each other. correlation, features extracted from the regions in question, between the images Determining their variability and regional availability, the information obtained converting regional feature reliability and visualizing that reliability differently 10 is used as weighting data in combining surface features obtained from the data. Its use is ensured. In addition, it is created with a focus on reliability. virtual makeup image created based on user-specific surface representation It is not considered a final and immutable outcome; the image in question is regional. The identified discrepancies and the factors influencing their formation are being re-analyzed. It is associated with the synthesis parameters and the generated correction data is related to the relevant 15 changing the values ​​of the synthesis parameters in the next synthesis cycle It is fed back in this way. Thus, multiple visual data processing and feedback virtualization are possible. makeup simulation system (1) to be used both before the synthesis process both evaluating the reliability of regional surface data and the synthesis process. then 20 functionally interoperable systems that check the regional consistency of the resulting image A connected image processing architecture is being created. Multiple visual data processing and feedback virtual makeup simulation system (1); Visual Data acquisition unit (2), Multiple visual dataset (3), Image processing unit (4), Data Storage unit (5), Face region determination unit (6), Skin region segmentation unit (7), Inter-region image matching unit (8), Image condition analysis unit (9), 25 Lighting feature determination unit (10), Exposure and shadowing determination unit (11), Regional image availability determination unit (12), Surface feature extraction unit (13), Inter-image feature variability determination unit (17), Regional feature Reliability determination unit (18), Reliability weighted feature combination unit (20), Stable surface feature creation unit (21), Visual context analysis unit (23), Clothing 30 Regional determination unit (24), Regional image synthesis parameter determination unit (28), Target lighting modeling unit (30), Virtual makeup image synthesis unit (32), Regional visual consistency inspection unit (34), Nonconformity classification and 16 Parameter matching unit (36), Parameter correction data due to mismatch creation unit (37), Feedback image synthesis parameter update unit (38), Consistency condition assessment and resynthesis control unit (39) and It includes a display and user interaction unit (40). These units It can be implemented as independent physical hardware elements, or 5 Software-based functional units that run on one or more processors or in the form of process structures where software and hardware components are used together This can be achieved. The functionally separate definition of the units is in question. The units are hardware components that are necessarily physically separated from each other. This does not mean... 10 The visual data acquisition unit (2) obtains at least two visual data relating to the same user. It ensures that the visual data is obtained by the visual data acquisition unit (2). Data from the same imaging device at different times and in different camera locations, different user poses, different viewing angles, or different lighting It can be obtained under these conditions, as well as from multiple imaging devices. 15 visual data obtained by the visual data acquisition unit (2) are the same. Multiple visual datasets (3) are determined to be user-related. It is related. Multiple visual dataset (3) two or more visual datasets relating to the same user. in addition to ample visual data, images that can be associated with that visual data time, image size, image orientation, image scale and viewing conditions 20 It may include at least one of the related supporting data. Multiple visual dataset (3) It is not necessary for the visual data within it to be obtained under the same viewing conditions, The variability between visual data obtained under different viewing conditions is the discovery. It constitutes one of the inputs of the reliability assessment carried out within this scope. It can bring. 25 Data within the multiple visual dataset (3) are processed by the Image processing unit (4). is processed. The image processing unit (4) performs subsequent regional comparison of the visual data. making the area suitable for the procedures, identifying the face and skin areas, different visual Matching corresponding regions in the data, regional surface features The removal of these features reduces the variability of these characteristics among different visual data. 30 determination, establishment of regional characteristic reliabilities, reliability-weighted performing feature merging, image synthesis process the most important of the processes related to execution and re-analysis of the synthesized image 17 It is structured to perform only a small part of it. Image processing unit (4), differences in size, orientation, scale, or coordinates between different visual data negatively impacting subsequent regional matching and comparison processes It can also perform image preprocessing operations to limit its capabilities. Data storage unit (5), Multiple visual data processing and feedback virtual makeup 5 The simulation system (1) is used, generated and between the processing cycles. It enables the storage of reusable data. Data storage unit (5) Multiple visual dataset (3), regional image data, extracted surface features, data on viewing conditions, regional availability information, reliability information, user-specific surface features, visual context data, image 10 synthesis parameters, target lighting information, synthesized images, mismatch values ​​and parameter correction generated in the feedback loop data can be stored. The data storage unit (5) can store the said data to the relevant user. The source image data will be associated with the face or skin region, feature type, and processing cycle. It can be structured in this way. Thus, 15 different processes are created at different stages. The data are correlated and provided with feedback in subsequent processing stages. This ensures that up-to-date data can be used in resynthesis cycles. Face region determination unit (6), visual within Multiple visual dataset (3) It identifies the user's facial region in the data. Facial region identification unit (6), facial boundaries, anatomical reference points within the face, or different sub-faces 20 It can determine the locations of the regions. Face region determination unit (6) The data generated reveals commonalities between user faces across different visual datasets. or allows processing within a regional coordinate structure that is convertible to one another. It provides. Skin region segmentation unit (7), Face region determination unit (6) determined by 25 separating areas of the face that have skin characteristics from other image areas. and divides these areas into multiple regional sub-areas when necessary. Volume The area is divided by the partitioning unit (7) into the forehead, cheek, nose area, chin, eye area and At least one of the similar facial regions is represented as a separate regional data space. It is possible to divide the entire face into a single 30-degree segment thanks to the skin region segmentation unit (7). Instead of being considered a homogeneous surface, it should be viewed as a surface with varying colors, tones, brightness, contrast, and face and skin regions that may have different surface appearance characteristics processing is ensured. 18 Inter-region image matching unit (8), Face region detection in different visual data The corresponding face determined through the unit (6) and the Skin region segmentation unit (7) and correlates skin regions with each other. Inter-regional image matching. unit (8), user's head position, camera angle, image scale, image orientation or Despite differences in facial expression, a particular face or skin tone found in a visual dataset is 5 This allows for the determination of the region's counterpart in other visual data. Thus, surface features extracted from different visual data can be compared. not between different anatomical regions, but between the same or corresponding regional regions. It is carried out between areas. Inter-regional image matching unit (8) The matching relationship created by, subsequent inter-image feature variability 10 by the determination unit (17) and the Regional characteristic reliability determination unit (18) It creates a regional comparison infrastructure for the processes to be carried out. The imaging condition analysis unit (9) analyzes each of the multiple visual datasets (3). It evaluates the display conditions of visual data. Display condition analysis. unit (9) provides general image quality information about the entire visual data only 15 not limited to creating, but also different facial and skin regions within the same visual data. regional It can perform an evaluation. In this way, the visual data in general Even if usable, shadowing of a particular face or skin area, reflection, overexposure or underexposure, chromatic aberration or similar imaging issues 20 It can also be determined that they are intensely affected by its effects. Lighting feature determination unit (10), within the Multiple visual dataset (3) It specifies the characteristics related to the lighting conditions under which visual data is obtained. Lighting characteristic determination unit (10), light intensity, regional illuminance, light directional distribution, color characteristics of light, or data representing at least one of these 25 It can be created by the lighting feature determination unit (10). data refers to the color, tone, or brightness characteristic extracted from a specific face or skin area. the level of representation of a stable surface characteristic belonging to the user It can be used in the evaluation. Exposure and shadowing determination unit (11), in different visual data and different faces or 30 overexposure, underexposure, and shadowing effects occurring in skin areas Exposure and shading determination unit (11) determines the effects in question. either in terms of the entire image or separately in terms of different facial and skin areas. 19 It is able to determine this. Thanks to this structure, it can be appropriately identified within the same visual data. Features extracted from an imaged region and intensive imaging conditions. the subsequent reliability of a feature extracted from another affected region This prevents the assessment from being deemed equivalent. Regional image availability determination unit (12), Image condition analysis 5 Unit (9), Lighting feature determination unit (10) and Exposure and shadow determination using at least one of the data generated by unit (11) different face or skin the usability of the regions in terms of surface feature extraction and feature blending It determines. The regional image availability determination unit (12) determines a visual It is not limited to creating a single accept or reject option for all the data, but also includes the same visual 10 Different usability information for different regions within the data. It can create. Thus, data obtained from a highly available region While ensuring protection in subsequent processing stages, it is intensive in imaging conditions. on a combined surface representation of data obtained from a region affected in this way It is possible to limit its effect. 15 Surface feature extraction unit (13), Skin region segmentation unit (7) determined and interregional image matching unit (8) different visual data regional surface features of regions that are correlated with their counterparts surface features extracted by surface feature extraction unit (13) Regional color feature data (14), Regional hue and brightness feature data (15) and 20 It includes regional contrast and surface appearance feature data (16). Regional color feature data (14) describes the color characteristic of the relevant face or skin region. It represents. Regional color feature data (14) consists of a single color value. not necessarily, but the distribution of color components in the relevant area, dominant color values, regional color differences or color representation data derived from them 25 It may include regional color feature data (14), from which the data in question was obtained. visual data and comparisons between images in relation to face or skin region, reliability assessment and reliability-weighted feature combining processes It is used. Regional tone and brightness feature data (15) indicates the tone and brightness of the relevant face or skin area. It represents brightness characteristics. Regional tone and brightness characteristic data. (15), light-dark distribution of the relevant region, regional brightness values, brightness It may include the distribution or feature data derived from them. Regional tone and brightness feature data (15) was obtained especially under different lighting and exposure conditions. Regional differences observed among the visual data obtained in subsequent images its use in assessing inter-trait variability and reliability It provides. Regional contrast and surface appearance feature data (16), 5 of the relevant face or skin region It represents the local contrast structure and visual surface characteristics. Regional contrast and surface appearance feature data (16), visual within the relevant region changes, local surface appearance and the visuals it creates with neighboring image areas It can contain data representing relationships. Thus, the virtual makeup image In its creation, not only color, tone and brightness information but also regional surface 10 It is also possible to use visual characteristics. Inter-image feature variability determination unit (17), Inter-region image obtained from facial or skin regions whose counterparts are determined by the matching unit (8) The obtained regional color feature data (14), regional hue and brightness feature data (15) and Regional contrast and surface appearance feature data (16) between images 15 It determines the variability. Inter-image feature variability determination unit (17), What is the corresponding regional feature of the same user across different visual data? It can generate data that represents whether something is stable or variable to a certain extent. Variability is considered solely as a physical or visual change unique to the user. not being done; Viewing condition analysis unit (9), Lighting feature determination unit 20 (10), Exposure and shadowing determination unit (11) and Regional image availability It is associated with the data generated by the determination unit (12). The regional feature reliability determination unit (18) corresponds to different visual data. features extracted from facial or skin areas reveal the user's surface characteristics. to determine its reliability in terms of representation and for the characteristics in question 25 Regional feature reliability value (19) is formed. Regional feature reliability Determination unit (18), Regional feature reliability value (19) are created while Images inter-image feature variability determined by the unit (17) by the Regional image availability determination unit (12) with variability They can jointly evaluate the determined regional availability information; also 30 Lighting features determined by the lighting feature determination unit (10) Exposure and shadowing determined by the exposure and shadowing determination unit (11) It can include at least one of the shading effects in the assessment. Thus... 21 The regional feature reliability value (19) reflects only the overall quality of the source image. from being information that represents something, to being information from a specific region in a particular visual data. To what extent does the extraction of a specific surface feature contribute to the subsequent data merging process? as regional and feature-based control data that determines what it will provide is being created. 5 Regional feature reliability determination unit (18), Regional feature reliability value (19) images relating to the same or corresponding face or skin area when created Increased inter-modulus feature variability, decreased regional image availability, lighting conditions negatively affect the determination of the relevant surface characteristic or At least one of the following situations occurs where exposure and shadowing effects increase on the relevant surface 10 It can be considered a data relationship that reduces the reliability level of the feature. In contrast, the corresponding regional surface feature is more pronounced across different visual data. low variability and better image availability of the region in question A high value is a data relationship that increases the reliability level of the relevant surface feature. It can be evaluated. Thus, the regional characteristic reliability value (19), 15 a dataset where independent evaluation results are stored together. by removing the inter-image feature variability and regional imaging. its availability will change the contribution of the relevant surface feature to the bonding process It is created as functional control data that is associated in this way. The regional feature reliability value (19) is in the form of a binary acceptance or rejection data 20 It can be created, as well as enabling the separation of different levels of reliability. It can also be created in the form of a gradual or continuous value. A specific face or a surface characteristic obtained from the skin region is high in corresponding areas The variability between images indicates the image availability of the region in question. 25 if it is low or heavily affected by imaging conditions The relevant Regional feature reliability value (19) of the surface feature in question is the next It can be formulated in a way that reduces its contribution to the combining process. In contrast... showing lower variability in corresponding regions in different visual data and more Regional characteristics relating to surface features obtained under suitable imaging conditions. The reliability value (19) is 30, which is the contribution of the features in question to the combined surface representation. It can be used in a way that will increase [its effectiveness]. Regional feature reliability value (19) is only the reliability of the relevant surface feature. It is not used as passive evaluation data that defines the situation, 22 Feature assembly performed by reliability weighted feature assembly unit (20) the merging process produces one of the numerical or graded weighting inputs It brings about different visuals relating to the same or corresponding face or skin area. Regional color feature data (14), Regional tone and brightness feature obtained from the data. data (15) and Regional contrast and surface appearance feature data (16) are 5 Different regional feature reliability values ​​(19) can be created and these values to determine the amount of contribution of the relevant feature to the combined regional surface representation It can be used. In this context, a higher Regional feature reliability value (19) the contribution of a surface feature associated with a combined surface feature While it can be increased, a 10 associated with a lower Regional feature reliability value (19) The contribution of surface feature can be reduced or a defined reliability condition can be minimized. If it falls below this threshold, the feature in question may be excluded from the merging process. Thus, temporary variability caused by imaging conditions is only detected. not only is this variability observed, but the surface characteristics in question are also specific to the end user. The effect on the data structure (22) is 15 through the regional feature reliability value (19). It is directly controlled. Reliability weighted feature combination unit (20), corresponding to different visual data Regional color feature data obtained from regions (14), Regional hue and brightness feature data (15) and Regional contrast and surface appearance feature data (16) are mentioned. Relationship between the Regional Feature Reliability Value (19) corresponding to the subject features 20 It performs a reliability-focused feature bonding process by establishing this. Reliability weighted feature combination unit (20), higher Regional feature Combined regional representation of surface features carrying reliability value (19) can increase its contribution to lower Regional feature reliability value (19) However, the contribution of surface features carrying the load can be limited. 25 The process performed by the reliability weighted feature combination unit (20), Surface extracted from different visual data within the multiple visual dataset (3) summing, directly selecting, or arithmetically adding up the features with equivalent weights It is not limited to centering. Inter-image feature variability determination unit (17) Regional variability and regional image availability determination as determined by 30 Regional availability determined by unit (12), Regional feature reliability Regional feature reliability value (19) by determination unit (18) It is used in its creation; the Regional feature reliability value (19) is 23 The relevant surface feature within the reliability weighted feature combination unit (20) This changes the level of contribution. In this way, it affects the imaging conditions and the inter-image relationship. Information regarding variability is considered passive data produced solely for evaluation purposes. not remaining, techniques that directly affect the formation of the combined surface representation It is converted into weighting data. 5 Combining performed by the reliability weighted feature combining unit (20) in the process, from different visual data relating to the same or corresponding face or skin area the contribution of each obtained surface feature to the combined regional surface feature contribution, Regional feature reliability value associated with the surface feature in question (19) can be determined separately based on different visual data. Different contribution levels to the same type of regional surface properties obtained. can be assigned and combined regional surface features are taken into consideration in these contribution levels. This can be done by taking into account the data. Thus, reliability information is only available at low levels. Not as a threshold information for filtering out data with low reliability, but as a different visual representation. The combined regional surface feature of surface characteristics obtained from the data is 15 as a weighting input that determines the extent to which it will affect its formation It is used. Stable surface feature creation unit (21), Reliability weighted feature combining By processing the reliability-weighted regional characteristics created by the unit (20), different User-specific surface characteristics that show continuity under viewing conditions 20 It forms. The stable surface feature formation unit (21) is formed from the imaging conditions. temporary changes in color, tone, brightness, contrast, or surface appearance resulting from limiting the impact of the user's surface representation and different visual data preserving the representation of regional characteristics observed more consistently throughout It provides. 25 The process performed by the stable surface feature formation unit (21) is different limited to creating a simple average of features obtained from visual data. is not. Stable surface feature creation unit (21), Reliability weighted feature Regional feature reliability value (19) is taken into consideration by the combining unit (20). It uses features that are combined from different visual data. Thus, 30 representation of regional surface characteristics that are repeated and have higher reliability while protected, it only appears under specific imaging conditions and is lower. 24 The impact of reliable temporary features on the generated user interface representation. It can be reduced. Regional created by the stable surface feature forming unit (21) The features are organized within the User-Specific Surface Feature Data Structure (22). User-specific surface feature data structure (22) categorizes different regions of the face into 5 associating reliability-weighted and stable surface characteristics with the regions It is a structured data representation. User-specific surface feature data structure (22) Each facial or skin region within it is accompanied by identifying information about that region. Regional color feature data (14), Regional hue and brightness feature data (15) and the top 10 derived from regional contrast and surface appearance feature data (16). It can be represented in such a way that a relationship is established between a few stable surface properties. User-specific surface feature data structure (22) allows direct access to a specific source visual data. a copy or an equally weighted average of features obtained from different visual data It is not of a nature. User-specific surface feature data structure (22), different visual Inter-image variability obtained from corresponding regions in the data is 15 identified, their regional availability assessed, and their reliability evaluated individually. as a result of the reliability-oriented processing of the determined surface properties It is a regional data structure created. User-specific surface feature data structure (22), different regions of the face independently or in relation to each other processing and determination of regional image synthesis parameters by the unit (28) 20 Data for creating regional image synthesis parameters (29) specific to the relevant regions It provides the infrastructure. Visual context analysis unit (23), in the synthesis of virtual makeup image The visual to be evaluated together with the user-specific surface feature data structure (22) It determines contextual features. Visual context analysis unit (23), Multiple visual data 25 image areas other than the face region within the cluster (3) or also obtained It can process visual context data. Visual context analysis unit (23) The data generated is only used to suggest a specific makeup look to the user or It is limited to use for selecting a predefined makeup option. not, but in the creation of regional parameters of the image synthesis process 30 The technical inputs used are being transformed into this data. Clothing area determination unit (24) identifies clothing within the user's visual data. It identifies one or more corresponding image regions. Clothing region Clothing color distribution from clothing areas determined by the determination unit (24) The clothing color distribution data (25) is generated. dominant colors, the distribution of color components, regional color relationships, or these It can include derived visual color representation data. Clothing color distribution data (25), 5 in contrast, in determining subsequent regional image synthesis parameters It is processed as usable visual contextual data. Environmental visual feature data (26), the user image is obtained or virtual makeup It represents the characteristics of the visual environment in which the image will be evaluated. Environmental visual feature data (26), background color distribution, ambient light-dark 10 in terms of its characteristics, environmental color effects, or image synthesis process. It may include at least one of the other available visual media characteristics. Media Visual feature data (26) must be the same as Target lighting data (31). not as a separate dataset for representing visual context It is available. 15 Visual context analysis unit (23), Clothing color distribution data (25), Environmental visual features by processing the relationship between the data (26) and the User-specific surface feature data structure (22) It forms a combined context feature data structure (27). Combined context feature data structure (27), stable regional surface features relating to the user and clothing and environment visual 20 that enables the combined evaluation of their features in terms of image synthesis It is a structured data representation. It is located within the combined context feature data structure (27). visual context features, a makeup category that will only be presented to the user, in choosing a cosmetic product or a pre-prepared makeup stencil User-specific surface features are not considered as supporting data; Regional image synthesis parameter determination unit (28) together with data structure (22) 25 by directly generating image synthesis parameters It is used. Unified context feature data structure (27), a makeover to be presented to the user only. a suggested data point for determining the option, cosmetic product, or visual preference It is not limited to use as. The combined context feature data structure (27) provides the User with 30 direct image synthesis process together with the specific surface feature data structure (22) This constitutes the technical input. In this context, user-specific surface feature data. Combined context with regional surface characteristics represented within the structure (22) 26 clothing and environmental features represented within the feature data structure (27) The established data relationship will be used for image synthesis applied to different regions of the face. It is used in determining behavior. Thus, in visual context data The change affects not only the option presented to the user, but also the Regional view. Regional image 5 created by the synthesis parameter determination unit (28) the value of at least one parameter within the synthesis parameters (29) or This can lead to changes in the way it is implemented. Regional image synthesis parameter determination unit (28), User-specific surface data in the feature data structure (22) and the combined context feature data structure (27) By processing, it creates regional image synthesis parameters (29). Regional 10 image synthesis parameters (29), virtual makeup image of different face color, tone, brightness, density, transparency, regional distribution and transitions in these regions It includes parameters that define at least one of its characteristics. Regional image Synthesis parameter determination unit (28), User-specific surface feature data structure (22) Combined context feature data with regional surface features represented within 15 by processing the relationship between the visual context features represented within the structure (27) Synthesizing different regional images for different facial or skin areas. It can create parameters (29). The regional image synthesis parameter determination unit (28) is specific to the same User. While the surface feature data structure (22) is preserved, the combined context feature data structure (27) 20 If the visual contextual characteristics represented within it change, then the aforementioned Regional view where the change is effective in terms of the relevant face or skin area. transferring at least one parameter value within the synthesis parameters (29) It can provide. Similarly, under the same visual contextual characteristics, to the User. 25 regional surface characteristics within the specific surface feature data structure (22) Differentiation, Regional image synthesis created for the relevant face or skin region. causing at least one parameter within the parameters (29) to be differentiated It is possible. Thus, the regional image synthesis parameters (29) are determined in advance. Instead of directly applying a defined, immutable makeup template, the user is given 30 Data relationship between specific regional surface features and visual context features It is created depending on the circumstances. Regional image synthesis parameters (29), single and applied to the whole face It does not have to be constructed in the form of a fixed set of parameters. Regional 27 Image synthesis parameter determination unit (28), e.g. cheek, forehead, eye area, For at least one of the following areas of the face: around the nose, chin, or other facial regions, the relevant area is provided to the User. surface characteristics within the specific surface feature data structure (22) and the Combined Separate Regional context features are based on visual context features within the context feature data structure (27). It can create image synthesis parameters (29). Thus, the same virtual makeup 5 color, tone, brightness, and intensity of different facial areas within its appearance, Transparency, distribution, or transition properties can vary regionally. Regional image synthesis parameters (29) are completely independent of each other. They can be created as regional values, as well as visual differences between neighboring facial regions. 10 in a way that is related to each other in order to protect the relationship It can be created. The regional image synthesis parameter determination unit (28) is a Regional image synthesis parameters (29) created for the face region and adjacent Regional image synthesis parameters (29) created for the face region It can evaluate color, tone, brightness, intensity, transparency, or transition relationships. Thus, the image synthesis parameter created for one region is 15 times that of the neighboring region. the subsequent situation of creating a sudden or unwanted visual discontinuity Determination and feedback by the regional visual consistency control unit (34) Infrastructure is provided for its correction within the system. Target lighting modeling unit (30) where the virtual makeup image will be created or Target lighting data (31) 20 representing the target lighting conditions to be evaluated It consists of target lighting data (31), target light intensity, direction of light incidence, light It may include color characteristics or data representing at least one of them. Target lighting modeling unit (30), the available lighting of the source visual data not limited to directly repeating the conditions but targeting a different goal than the source visual data. The use of lighting conditions in creating a virtual makeup look. 25 It provides an opportunity. Target lighting data (31) is added to a completed virtual makeover image afterwards. It is not limited to being used as a general brightness or color transformation. Target illumination data (31), together with regional image synthesis parameters (29). The image synthesis process is 30 by the virtual makeup image synthesis unit (32). It is used as one of the inputs. User-specific surface feature data structure (22) different facial regions represented within it under the same target lighting conditions Target lighting data (31) can show different visual responses. 28 Regional image synthesis parameters (29) for different facial regions This can be associated with target illumination on a regional surface. its effect on characteristics, after the virtual makeup image is created instead of a uniform image effect applied, within the image synthesis process It can be evaluated. 5 Target illumination data (31) and regional image synthesis parameters (29) The relationship between them is that the target lighting condition is equivalent to the entire face. It is not limited to its application. User-specific surface feature data structure (22) The same target lighting is used, taking into account the represented regional surface characteristics. data (31) different image synthesis results occur in different face or skin regions 10 Regional image synthesis parameters (29) in such a way as to bring It can be associated. Similarly, User-specific surface feature data structure (22) and Regional image synthesis parameters (29) are preserved while a different Target The use of lighting data (31) will create the color in the relevant face or skin areas. at least one of the following characteristics: tone, brightness, density, or surface appearance is a target of 15 This can cause variations depending on the lighting conditions. Thus Target illumination data (31) subsequently alters the entire generated image. instead of a general image transformation, the result of a regional image synthesis process. It is used as a synthesis input that directly affects the process. Virtual makeup image synthesis unit (32), User-specific surface feature data 20 structure (22), Regional image synthesis parameters (29) and Target illumination data (31) creates a synthesized virtual makeup image (33). Virtual Makeup image synthesis unit (32), Regional corresponding to different regions of the face color, tone, brightness, density in accordance with image synthesis parameters (29), It synthesizes at least one of the features of transparency, regional distribution, and transition, and the word 25 The subject is represented by the Target illumination data (31) during the synthesis process It takes into account the target lighting conditions. Synthesized virtual makeup image (33), User-specific surface feature data structure (22) Stable surface features represented in the Unified context feature data structure Visual context features represented by (27), Regional image synthesis 30 Digital virtual created using parameters (29) and target lighting data (31). It is a makeup image. Synthesized virtual makeup image (33), Multiple visual data direct final processing and feedback virtual makeup simulation system (1) 29 and is not considered an immutable outcome, but regional after its creation Regional visual consistency audit unit for the purpose of evaluating visual consistency. (34) is being reprocessed by. Regional visual consistency control unit (34), Synthesized virtual makeup image (33) re-analyzes and synthesizes different facial or skin regions within it. 5 The resulting regional image features are user-specific surface feature data. structure (22), Regional image synthesis parameters (29), Target illumination data (31) and its consistency with at least one of the image features in adjacent facial regions The regional visual consistency control unit (34) evaluates the matter. in the evaluation color, tone, brightness, intensity, transparency, regional distribution or 10 It can utilize at least one of the regional transition features. In the assessment carried out by the regional visual consistency control unit (34), from the relevant face or skin area in the synthesized virtual makeup image (33) Image synthesis with respect to the region in question using the redefined image feature. Among the regional reference image features determined from the data used in the process are 15. A comparative relationship can be established. This regional reference image feature, User-specific surface feature data structure (22) associated with the relevant region surface features, regional image synthesis parameters used for the relevant region (29), Target illumination data (31) and, where necessary, adjacent facial or skin areas. It can be determined using at least one of the image features. Regional visual 20 consistency control unit (34), thus Synthesized virtual makeup image (33) not limited to evaluating the overall image quality within, but also a specific face or synthesis with the regional reference image feature in the skin region As a result, it can identify the difference between the re-determined image feature. The identified differences include color, tone, brightness, density, transparency, regional distribution, or 25 quantitatively or progressively expressed in terms of at least one of the regional transition characteristics. It can be done. Thus, the regional visual incongruity value (35) is a subjective visual not as an evaluation-based result, but an image of a specific face or skin area. Regional reference image feature determined from data used in the synthesis process. Synthesized virtual makeup image (33) redefined image 30 as technical control data representing the difference between the features can be created. As a result of the evaluation carried out by the regional visual consistency audit unit (34), Regional visual dissonance value (35) is formed. Regional visual dissonance The value (35) is a binary control information indicating whether or not there is a mismatch. It does not have to be created as such. Regional visual incompatibility value (35), the face or skin area where the discrepancy occurred, the 5 areas associated with the discrepancy in a way that represents the image characteristic and the extent of the discrepancy in question It can be created. Thus, the synthesized virtual makeup image (33) Which region has which image feature and at what level of correction? It can be determined what is required. Regional visual discordance value (35), Target 10 in a specific face or skin area The difference between the expected and synthesized brightness in terms of lighting data (31), The regional feature represented within the user-specific surface feature data structure (22) Synthesized virtual makeup image (33) from the redefined regional the difference between the features, the discontinuity of color or tone between adjacent facial regions, 15 Thus, the Regional visual consistency control unit (34) can do so. The evaluation conducted is based solely on user preference or subjective aesthetics. It is not an evaluation-based control but is used in the image synthesis process. the relationship between technical data and image characteristics that can be determined from the synthesis result. It is carried out as a regional technical assessment based on the relationship. 20 Incompatibility classification and parameter matching unit (36), Regional visual classifying the discrepancy represented by the discrepancy value (35) and the said discrepancy at least one of the related Regional image synthesis parameters (29) is incompatible. It matches the parameter. Incompatibility classification and parameter matching unit. (36), the region where the discrepancy occurred, image 25 with which the discrepancy is associated the feature and the Regional used in creating that image feature the parameter or parameters together within the image synthesis parameters (29) is evaluating. Performed by the incompatibility classification and parameter matching unit (36) matching, any Regional image with Regional visual discrepancy value (35) 30 in the form of random association of the synthesis parameters (29) element is not being implemented. Regional visual discordance value (35) for a specific face or skin when it represents color disharmony in a region, it refers to the color formation of that region. 31 The relevant parameter among the regional image synthesis parameters (29) that affect, When representing brightness mismatch, the relevant parameter affecting brightness generation, when it represents a density or transparency mismatch, the relevant density or when the transparency parameter represents the discontinuity of the transition between neighboring regions The relevant parameter influencing the occurrence of the transition in question can be determined. Thus, 5 A measurable regional area is identified within the synthesized virtual makeup image (33). image inconsistency and factors contributing to the occurrence of that inconsistency A functional relationship is established between the image synthesis parameters. Parameter correction data generation unit due to incompatibility (37), Regional visual Incompatibility value (35) and Incompatibility classification and parameter matching unit (36) 10 Using the parameter matching information generated by, the relevant regional image It generates correction data for synthesis parameters (29). Mismatch The source parameter correction data generation unit (37), where the mismatch occurs the region of origin, the type of incompatibility, and the amount of incompatibility together by evaluating which Regional image synthesis parameters (29) element which 15 data representing the direction or amount of change required. It can create. by the unit for generating parameter correction data due to incompatibility (37) The generated correction data is based on the facial or skin area to which the correction will be applied. Regional image synthesis parameters (29) 20 associated with correction of the subject in a way that will create a regional and parameter-specific relationship between the parameters within it It can be configured. In this context, a specific facial or skin area is determined The regional visual inconsistency value (35) is the regional visual inconsistency value for the entire virtual makeup image. Changes occur indiscriminately on the image synthesis parameters (29). Instead of bringing, Incompatibility classification and parameter matching unit (36) 25 At least one Regional image associated with the aforementioned discrepancy. selective modification of the synthesis parameters (29) element It can be converted into correction data. Thus, the discrepancy occurred. region, type and amount of non-conformity and the non-conformity in question Among the image synthesis parameters to be modified in order to reduce this, there are 30 direct changes. A technical data link is being established. Parameter correction data generation unit due to incompatibility (37), Regional visual the direction of the relationship between the mismatch value (35) and the targeted regional situation and 32 Regional image synthesis parameters depending on its size (29) increasing, decreasing, or changing the value of a parameter within it It can generate correction data regarding the redirection. This correction data... an absolute new parameter value, the change to be applied to the existing parameter value the amount, a correction factor or 5 to be used in the next synthesis cycle in the form of equivalent control data that allows the parameter value to be determined It can be created. Thus, the Regional visual discrepancy value (35) is only detected. It does not remain as an error or warning data point, but is used in subsequent image synthesis. in generating active control data that changes the parameters of the cycle It is used. 10 Feedback image synthesis parameter update unit (38), Mismatch Correction generated by the source parameter correction data generation unit (37) Depending on the data, the most relevant Regional image synthesis parameters (29) It updates a small parameter. Feedback image synthesis parameter. Update performed by update unit (38), Regional visual 15 tiered depending on the amount of discrepancy represented by the discrepancy value (35) a parameter change, the current parameter value and the correction data together evaluation or reduction of the identified discrepancy in the relevant parameter This can be accomplished in the form of an equivalent parameter update operation. The feedback image synthesis parameter update unit (38) by 20 Updated Regional image synthesis parameters (29), re-virtual makeup The virtual makeup image is transferred into the synthesis unit (32). synthesis unit (32), updated Regional image synthesis parameters (29), User-specific surface feature data structure (22) and Target lighting data (31) It creates a new Synthesized virtual makeup image (33) using 25 Thus, the regional areas determined on the previous Synthesized virtual makeup image (33) technical incompatibility directly affects the parameters of the subsequent image synthesis process. It is converted into feedback data that modifies the process. Consistency condition assessment and resynthesis control unit (39), updated The current 30 created using regional image synthesis parameters (29) Regional visual consistency check on synthesized virtual makeup image (33) Current Regional visual mismatch value (35) determined by unit (34) or 33 a consistency defined through multiple regional visual discrepancy values ​​(35) It assesses whether the condition is met or not. Consistency condition assessment and resynthesis control unit (39) The consistency condition being evaluated is not limited to a single fixed threshold value. Different faces or skin regions, different types of discrepancies, or different regional image synthesis 5 Different acceptance conditions can be defined for the parameters (29); more than one A combined consistency is assessed by evaluating the regional visual inconsistency value (35) together. The condition can be established. Consistency condition evaluation and resynthesis. When the control unit (39) determines that the said consistency condition is not met Regional visual consistency control unit (34), Nonconformity classification and parameter 10 Matching unit (36), Incompatibility-related parameter correction data generation unit (37), Feedback image synthesis parameter update unit (38) and Virtual feedback processing performed between the makeup image synthesis unit (32) It enables the cycle to be re-run. The updated synthesized virtual makeup is applied in each iteration of the feedback process cycle. image (33) re-evaluated by the Regional visual consistency control unit (34) is being evaluated, current Regional visual incongruity value (35) is being created, word Subject of incompatibility Incompatibility classification and parameter matching unit (36) It is matched with the relevant Regional image synthesis parameters (29), Parameter correction data generation unit (37) created new 20 Correction data is being generated, Feedback image synthesis parameter. The relevant Regional image synthesis parameters (29) by the update unit (38) is being updated and new by the Virtual makeup image synthesis unit (32) A synthesized virtual makeup image (33) is created. The said cycle, Consistency condition assessment and resynthesis control unit (39) by 25 until a specified consistency condition is met or a defined transaction control limit is reached. It can be continued until the desired result is achieved. Thus, the resynthesis process... This prevents it from continuing uncontrollably and ensures closed-loop image synthesis. The process is carried out in accordance with defined technical control conditions. Consistency condition assessment and resynthesis control unit (39), sequential back 30 Regional feeding cycles are created for the same facial or skin area. It can also evaluate the change in the visual discrepancy value (35) data. Regional visual mismatch identified in the previous synthesis cycle within this scope 34 Regional visual discordance determined in the current synthesis cycle with value (35). Change between the value (35) and the relevant Regional image synthesis parameters (29) the impact of the update performed on regional visual inconsistencies It can be used in determining the current regional visual discrepancy value (35). when it meets the consistency requirement defined for the relevant face or skin area, the 5 in question Parameter correction and resynthesis process in terms of region. can be terminated; the relevant non-conformity if the consistency condition is not met. Re-classification and parameter matching unit (36), Incompatibility Source parameter correction data generation unit (37) and Feedback image The synthesis parameter can be processed via the update unit (38). Thus, 10 The continuation of the feedback resynthesis cycle is only possible with predefined parameters. not necessarily tied to a fixed number of repetitions, current regional inconsistencies depending on the situation and the relationship of the discrepancy in question to the defined consistency requirement It can be controlled as follows. Consistency condition assessment and resynthesis control unit (39) by 15 If it is determined that the consistency condition is met, the current Synthesized virtual makeup image (33), through the Display and user interaction unit (40) is presented to the user. Display and user interaction unit (40), Synthesized It provides visual interaction between the virtual makeup image (33) and the user and It allows the user to view the created virtual makeup look. 20 The display and user interaction unit (40) displays different target lighting conditions. It can also allow for the selection of a different target lighting condition. In this case, the Target lighting modeling unit (30) will create a new Target lighting Virtual data (31) can be generated and new Target lighting data (31) can be used Makeup image synthesis unit (32), Regional visual consistency control unit (34), 25 Feedback image synthesis parameter update unit (38) and Consistency condition assessment and resynthesis control unit (39) The relevant transactions that have been carried out can be repeated. Multiple visual data processing and feedback virtual makeup simulation system (1) In the first technical data processing chain carried out, the Visual data acquisition unit (2) 30 The visual data obtained and correlated within the Multiple visual dataset (3), Face region determination unit (6), Skin region segmentation unit (7) and Interregional view It is regionally associated via the matching unit (8); Display condition 35 Analysis unit (9), Lighting feature determination unit (10), Exposure and shadowing Determination unit (11) and Regional image availability determination unit (12) through the viewing conditions and availability of the regions in question Regional color feature data is determined via the Surface feature extraction unit (13). (14), Regional tone and brightness feature data (15) and Regional contrast and surface appearance 5 Feature data (16) is generated; Inter-image feature variability determination unit (17) corresponds to regional features between different visual data. Variability is determined; Regional feature reliability determination unit (18) Regional feature reliability value (19) is created through; Reliability weighted regional 10 obtained from different visual data through feature combining unit (20) The features are combined in accordance with the Regional feature reliability value (19) and User-specific surface properties through the stable surface property creation unit (21) data structure (22) is created. In the first technical data processing chain, the Interregional image matching unit (8), Inter-image feature variability determination unit (17), Regional feature reliability 15 Determination unit (18), Regional feature reliability value (19), Reliability weighted Feature Combining Unit (20), Stable Surface Feature Creation Unit (21) and User Thanks to the functional relationship established between the specific surface feature data structure (22) multiple visual Data is not only used as a resource to increase the amount of data. Different visual The surface features obtained from the corresponding regions in the data are 20% between the images. variability and regional availability, combined user experience of these features Regional feature reliability determines the extent to which it will contribute to surface representation. It is used in the creation of the value (19). Thus, different display User-specific surface heavily influenced by regional characteristics and conditions. While the impact on the feature data structure (22) is limited, across different visual data 25 the influence of regional characteristics identified as more stable and more reliable It can be increased. Multiple visual data processing and feedback virtual makeup simulation system (1) In the second technical data processing chain performed, user-specific surface feature data was obtained. structure (22), Visual context analysis unit (23) by Clothing area determination unit 30 Clothing color distribution data (25) and Environment visual feature data determined through (24). (26) is evaluated together with the Unified context feature data structure (27) It is created. Regional image synthesis parameter determination unit (28), 36 User-specific surface feature data structure (22) and Unified context feature data structure (27) Regional image synthesis parameters (29) by processing the relationship between them to create; Target lighting by Target lighting modeling unit (30) data (31) is generated and the Virtual makeup image synthesis unit (32), User-specific surface feature data structure (22), Regional image synthesis 5 Synthesized virtual makeup using parameters (29) and target lighting data (31). It forms the image (33). In the second technical data processing chain, Clothing color distribution data (25) and Environment visual feature data (26), only a makeover option to be presented to the user It does not remain as suggested data used in the determination, but rather as combined context feature data 10. Regional image synthesis parameter determination unit (28) via structure (27) Directly by the creation of Regional image synthesis parameters (29) It is used. Thus, the influence of visual context on a choice or classification not limited to the result of the Synthesized virtual makeup image (33) at least 15 regional color, tone, brightness, density, transparency, distribution or transition characteristics It is transferred to the formation of a small one. Similarly, Target illumination data (31), as a standalone visual effect performed on the finished image not limited to use by the Virtual makeup image synthesis unit (32) By evaluating together with regional image synthesis parameters (29), the target The influence of lighting on regional synthesis is included in the image formation stage. 20 It ensures that it is done. Multiple visual data processing and feedback virtual makeup simulation system (1) Synthesized virtual makeup was performed in the third technical data processing chain. The image (33) was reanalyzed by the Regional visual consistency control unit (34). Regional visual incongruity value (35) is created; Incongruity 25 classification and parameter matching unit (36) in question Regional visual mismatch The functional relationship between the regional image synthesis parameters (29) and the value (35). establishing a relationship; Parameter correction data generation unit due to incompatibility (37) parameter correction data depending on the relationship in question and the amount of discrepancy It creates; Feedback image synthesis parameter update unit (38) 30 Update on the relevant Regional image synthesis parameters (29) It performs and updates the Virtual makeup image synthesis unit (32). New Synthesized virtual makeup with regional image synthesis parameters (29) 37 It forms the image (33). Consistency condition evaluation and resynthesis control unit (39), depending on the defined consistency condition of the cycle in question It is checking whether to continue or not. The third technical data processing chain in question is the synthesized virtual makeup image. (33) is not only used as an image output presented to the user, but also 5 Regional image synthesis parameters (29) that make up the time a technique that gives rise to changes in the next processing cycle It is being turned into a data source. Regional visual consistency control unit (34) Regional visual incongruity value (35) generated by, Incongruity The mismatch occurs through the classification and parameter matching unit (36) 10 Regional image synthesis parameters (29) which are effective in its arrival It is associated with and the unit for generating parameter correction data due to mismatch. (37) parameter correction data aimed at reducing the said discrepancy is converted. In this way, the result of the synthesis and the subsequent synthesis cycle are transformed. A direct feedback control relationship is formed between the parameters. 15 is being brought in. Multiple visual data processing and feedback virtual makeup simulation system (1) mutually agreeing on the reliability of the input data and the regional consistency of the synthesis output Two complementary technical audit layers are created. The first technical audit Inter-region image matching unit (8), Inter-image feature 20 in the layer Unit for determining variability (17), Unit for determining regional characteristic reliability (18), Regional feature reliability value (19), Reliability weighted feature combination unit (20), Stable surface feature creation unit (21) and User-specific surface feature data regional features obtained from different imaging conditions through the structure (22) 25 Regional visual consistency control is provided in the second technical inspection layer. unit (34), Regional visual incongruity value (35), Incongruity classification and Parameter matching unit (36), Parameter correction data due to mismatch creation unit (37), Feedback image synthesis parameter update unit (38) and Consistency condition evaluation and resynthesis control unit (39) 30 the related regional incompatibilities that occur as a result of synthesis through Regional image synthesis parameters (29) are transferred back onto it. 38 These two technical control layers are independent blocks of operation. It is not limited to work. Reliability is established in the first technical audit layer. Predominant regional features through the Stable surface feature creation unit (21) The user-specific surface feature data structure (22) is structured within the user; Specific surface feature data structure (22), Regional image synthesis parameter determination 5 In the creation of regional image synthesis parameters (29) by unit (28). is used; the Regional image synthesis parameters (29) and Virtual Synthesized makeup image synthesizer unit (32) virtual makeup image (33) reanalyzed in second technical inspection layer Regional visual discrepancies identified in the second technical inspection layer are 10. The value (35) is matched with the relevant Regional image synthesis parameters (29) and The same through the feedback image synthesis parameter update unit (38) This leads to changes in the synthesis parameters in the next cycle. Thus, by determining the reliability of the input data, the regional synthesis result is obtained. The consistency of the images is evaluated within the same image synthesis architecture, 15 of which are interconnected. It is connected. Multiple visual data processing and feedback virtual makeup simulation system (1) The data flow created within the scope of this is therefore only the Visual data acquisition unit (2) In terms of the Viewing and User Interaction Unit (40) a linear progression It does not form an image processing chain. Visual data acquisition unit (2) 20 The reliability of the data obtained by regional image synthesis before image synthesis It is evaluated as follows: reliability information, reliability-weighted feature combining. The unit (20) is transferred to the creation of the user-specific surface representation and The synthesized virtual makeup image (33) created from this representation is re-analyzed. Regional discrepancy information obtained by feedback image synthesis 25 Regional image synthesis via parameter update unit (38) The parameters (29) are transferred back onto the system. In this way, the data on the input side of the system Reliability assessment and output-side synthesis result assessment, Closed-loop interconnected via regional image synthesis parameters (29) It creates an image processing structure. 30 By the multi-visual data processing and feedback virtual makeup simulation system (1) resulting from different imaging conditions due to the procedures performed regional variations in color, tone, brightness, contrast, and surface appearance are presented to the user. 39 limiting the transfer of equivalent weight to the specific surface representation; display conditions, regional availability and inter-image feature variability User-specific surface feature data of surface properties determined as more reliable Increasing the contribution of the structure (22); Clothing color distribution data (25) and Environment Regional image 5 of the visual context represented by visual feature data (26). Synthesis parameters (29) are used in creating; Target lighting data Target lighting conditions represented by (31) to the image synthesis stage inclusion of the regional elements in the synthesized virtual makeup image (33) determining the location, type and quantity of the discrepancies and the issues in question 10 The discrepancies were controlled on the relevant Regional image synthesis parameters (29) The data is then converted into feedback that causes changes. Thus, a multi-visual data processing and feedback virtual makeup simulation system. (1) Inter-region image matching unit (8), Inter-image feature Unit for determining variability (17), Unit for determining regional characteristic reliability (18), Regional feature reliability value (19), Reliability weighted feature combination unit 15 (20), Stable surface feature creation unit (21) and User-specific surface feature data Reliability-based multi-visual data processing structure created with structure (22); To the User Specific surface feature data structure (22), Combined context feature data structure (27), Regional Image synthesis parameter determination unit (28), Regional image synthesis parameters (29), Target illumination modeling unit (30), Target illumination data 20 (31) and regional image created with the Virtual makeup image synthesis unit (32) Synthesis structure; Regional visual consistency control unit (34), Regional visual mismatch value (35), Mismatch classification and parameter matching unit (36), Data generation unit for parameter correction due to mismatch (37), Feedback Image synthesis parameter update unit (38) and Consistency condition 25 feedback created with the evaluation and resynthesis control unit (39) Image correction structures are functionally interconnected within the same technical architecture. They are related. Within this integrated structure, regional surfaces are obtained from multiple visual data. determining the reliability of the characteristics and the reliability of said characteristic 30 determined as a result of synthesis by using it in weighting the fusion. regional mismatches are matched with the relevant synthesis parameters. Converting it to corrective data involves complementary technical processes. 40 The first process involves transferring user-specific surface properties to the image synthesis process. data structure (22) resulting from different display conditions of the data within it the second process ensures that it is created in a more stable manner against variations. Measurable regional in the synthesized virtual makeup image (33) Discrepancies in the subsequent synthesis cycle Regional image synthesis 5 This enables targeted changes to occur on the parameters (29). technical control of both input and output data of the image synthesis process. It is kept under control. Multiple visual data processing and feedback virtual makeup simulation system (1), this Thanks to its integrated technical structure, it only identifies facial features and offers 10 makeup options. a system that creates or displays a virtual makeup look to the user not; it involves matching corresponding facial and skin regions in different visual data, word The subject is inter-image variability and regional variation of features extracted from the regions. Determining reliability based on usability information, reliability Using the values ​​as weighting data in the feature combining process, 15 User-specific surface feature data structure (22) from reliability-weighted features by creating and processing the visual context features of the data structure in question. Creation of regional image synthesis parameters (29), Target illumination data (31) Synthesized virtual makeup image (33) is created under, synthesized Regional discrepancies in the image are expressed as Regional visual discrepancy value (35) 20 Identifying these discrepancies in the relevant Regional image synthesis matching with parameters (29), discrepancies in parameter correction data by transforming and updating the relevant parameters through feedback. the virtual makeup image is recreated depending on the specified consistency condition. This creates an integrated and closed-loop image processing architecture in which the two components are synthesized. It brings. Industrial Application of the Invention: The invention relates to computer-aided image processing, computer vision, digital image synthesis, and virtual reality. appearance creation, augmented reality, e-commerce, cosmetic technologies and 30 in industry sectors where personalized digital user applications are developed It is manufacturable and applicable. The invention may be realized using a camera or... similar imaging equipment, programmable data processing equipment, data storage media, electronic imaging systems, user interfaces, and 41 Communication infrastructures can be used. The functional structures that make up the system. software-based modules, hardware-based components, or software and hardware Because it can be implemented in integrated structures where its components are used together, The application of the invention is not limited to a specific electronic device architecture. The invention relates to personal computers, portable computers, smartphones, tablet computers, 5 smart mirrors, cosmetic product sales and display terminals, in-store digital display systems, virtual trial systems, e-commerce platforms, online cosmetics applications and similar electronic systems with image processing capabilities It is possible to obtain and process visual data using the same electronic device. as can be performed on it, image acquisition, data processing, data storage, image 10 synthesis and presentation processes to the user using different electronic devices or remote data It is also possible to distribute it among processing infrastructures. Within the scope of the application of the invention, different viewing options for the same user and Multiple visual data obtained under different lighting conditions can be processed together. By identifying the corresponding facial and skin regions within different visual data, the statement 15 the color, tone, brightness, contrast and surface appearance characteristics of the regions in question can be extracted. Lighting, exposure, shadowing, and other factors used to obtain visual data, Image processing of different facial and skin regions with reflection and similar imaging conditions regional features extracted by evaluating their usability in this regard Their reliability can be determined. 20 Features extracted from corresponding facial and skin regions in different visual data The variability between the images can be determined, and that variability, by evaluating viewing conditions and regional availability information Regional feature reliability values ​​can be generated. These are obtained from different visual data. The surface characteristics obtained are weighted at 25 depending on the reliability values ​​in question. They can be combined, thus being heavily influenced by imaging conditions. More reliable while limiting the impact of features on the generated surface representation. And the contribution of more stable features can be increased across different images. This process resulting in temporary issues specific to the user and arising from varying viewing conditions. A regional surface feature data structure 30 in which the effect of visual variability is reduced can be created. Within the scope of the invention, user-specific surface characteristics are derived from the user's clothing areas. with the obtained color distribution information and visual characteristics of the display environment 42 They can be processed together. The data in question is only relevant to a specific makeup option or... not limited to recommending cosmetic products to the user, virtual makeup an image that determines how the image will be created in different facial regions as technical input in generating synthesis parameters It can be used. In this context, color, tone, brightness, 5 for different face and skin areas. density, transparency, regional distribution, and transition characteristics that specify at least one of them Different image synthesis parameters can be created. The lighting conditions targeted in creating the virtual makeup look are also important for the image. It can be included in the synthesis process. The light intensity related to the target illumination, the light Data such as direction of arrival and color characteristics are generated, and this data is then categorized into regional 10 image synthesis parameters and user-specific surface properties It can be processed. Thus, the virtual makeup look created is only from the source. Instead of the image remaining dependent on the existing lighting conditions, different uses or synthesis under target illumination conditions representing imaging environments It is possible. 15 The virtual makeup image created within the scope of the invention is a final and unchanging image. It does not have to be used as the output. Created after the synthesis process. The image is re-analyzed to determine the color, tone, and brightness of different facial and skin areas. Regional visual consistency in terms of density, transparency, distribution, and transition properties. The control can be performed. The synthesized image is surface-specific to the user. 20 with its characteristics, target lighting conditions and relevant regional synthesis conditions The discrepancies that arise as a result of the comparison are in terms of region, type, and quantity. This can be determined in terms of... The identified regional visual inconsistencies indicate the cause of the inconsistency. 25 that are effective in causing it or that need to be changed in order to reduce incompatibility It can be correlated with image synthesis parameters. The region of discrepancy, Depending on the type and quantity, the relevant image synthesis parameter is modified. Correction data can be generated for this purpose, and this correction data can be displayed on the image. This information can be fed back into the synthesis process. Thus, the relevant image is synthesized. By updating at least one of its parameters, the virtual makeup look will be restored to 30. can be created. Regional visual consistency check on the recreated virtual makeup image. It can be repeated, and the current discrepancy values ​​are determined in accordance with the established consistency. 43 It can be assessed whether the condition is met. The consistency condition... If this is not provided, discrepancy detection and image synthesis related to the discrepancy. matching with the parameter, generating parameter correction data, the relevant parameter update and resynthesize virtual makeup image processes It can be repeated. Thus, the virtual image synthesis process is only required for input 5. By moving away from being a one-way process progressing from data to image output feedback-based and closed-loop synthesis where the synthesis result modifies the next synthesis cycle. This can be accomplished within a cyclical image processing structure. The invention is particularly relevant to systems for digitally testing cosmetic products. virtual makeup applications, e-commerce platforms, in-store cosmetics 10 in testing terminals, smart mirror systems, mobile cosmetic applications, Personalized digital appearance in online cosmetics sales and promotion systems. simulation of user view in applications and under different lighting conditions It can be used in image processing systems where the desired outcome is sought. The user, 15 virtual appearance created before performing a physical cosmetic procedure under different visual contexts and different target lighting conditions can evaluate. Image acquisition, data processing, and data storage used to realize the invention, Image synthesis and electronic imaging components can be manufactured industrially. It is programmable and interconnected. The functional structures of the system are 20 software modules that can be loaded onto electronic devices, server-based processing structures, distributed data processing infrastructures or integrated systems specific to particular use cases This can be realized in the form of electronic devices. Therefore, the invention can be implemented in series. It is capable of being produced, reproduced, and applied to different electronic platforms. An invention is not merely a theoretical assessment or an abstract proposal approach. 25 not providing; regional visual data obtained with physical imaging tools processing as, surface extracted from corresponding regions in different images determining the reliability of the characteristics, the reliability of the characteristics in question its use in the merging process, deriving user-specific data from reliability-weighted data. Creating a surface representation, visual context and target lighting in relation to that representation 30 Generating regional image synthesis parameters from data, virtual makeup synthesis of the image, regional visual representation resulting from the synthesis quantitative determination of discrepancies and the creation of structures from the identified discrepancies 44 by feeding correction data back to the relevant image synthesis parameters This demonstrates a technically sound and repeatable image processing system. In terms of its characteristics, the invention relates to computer-aided image processing and cosmetic technologies. virtual view simulation, e-commerce and digital user interaction systems It is possible to apply it in industrial fields for its development and production. 5

Claims

45 REQUESTS 1. Data obtained for the same user under different viewing and lighting conditions. Virtual makeup look created by processing multiple visual data. A multi-visual data processing and feedback virtual makeover aimed at its creation. simulation system (1) and its feature is; different viewing for the same user and 5 obtaining multiple visual data under different lighting conditions Visual data acquisition unit (2) providing; faces in the said visual data Face region determination unit (6) which enables the determination of the regions; determined Skin region, which provides the segmentation of skin regions within facial areas. partitioning unit (7); corresponding face and skin in different visual data 10 Inter-region image matching unit (8) which matches the regions; matched face and Surface enables the extraction of regional surface features from skin areas. feature extraction unit (13); extracted from the corresponding face or skin regions determining the variability of surface properties between different visual data Inter-image feature variability determination unit (17); the images in question are 15 inter-feature variability and regional appearance relating to the relevant facial or skin areas. Regional for surface features extracted by jointly evaluating their usability. Regional feature reliability determination which forms feature reliability value (19) unit (18); from the corresponding face or skin regions in different visual data the contribution of the extracted surface features to the combined regional surface feature is mentioned 20 The subject depends on the Regional characteristic reliability value (19) Reliability weighted feature combination unit (20); reliability User-specific surface from predominantly combined regional surface features Stable surface feature creation which enables the creation of feature data structure (22) unit (21); visual 25 with the User-specific surface feature data structure (22). by considering contextual features together for different face or skin areas Regional image synthesis parameters (29) forming the Regional image Synthesis parameter determination unit (28); target lighting condition Target lighting modeling that represents Target lighting data (31) unit (30); User-specific surface feature data structure (22), Regional image 30 using synthesis parameters (29) and target illumination data (31) Synthesized virtual makeup image (33) Virtual makeup image Synthesis unit (32); Face in synthesized virtual makeup image (33) 46 or image synthesis with redefined image features in skin areas Regional reference image characteristics determined from the data used in the process. by evaluating the difference between them, the relevant face or skin area, image feature Regional visual discrepancy value (35) representing the type and amount of difference Regional visual consistency control unit (34) which ensures the creation of the aforementioned 5 The mismatch represented by the regional visual mismatch value (35) Regional classifications need to be changed in order to reduce discrepancies. matching with at least one parameter in image synthesis parameters (29) Incompatibility classification and parameter matching unit (36); the said Based on the matching and regional visual discordance value (35), the relevant face or 10 skin region and related Regional image synthesis parameters (29) Mismatch-related parameter that generates correction data for the parameter. Correction data generation unit (37); depending on the generated correction data At least one of the relevant Regional image synthesis parameters (29) Regional image synthesis parameters (29) 15 The data is synthesized by the Virtual makeup image synthesis unit (32) enabling its use in the subsequent image synthesis process Feedback image synthesis parameter update unit (38) and using updated regional image synthesis parameters (29) The current Synthesized virtual makeup image (33) created in terms of Regional 20 visual inconsistency value (35) satisfies the defined consistency condition of the data by assessing that it does not provide it, and when the consistency condition in question is not met mismatch detection, parameter matching, parameter correction, parameter Consistency enables the repetition of update and resynthesis processes. 25 by including condition assessment and resynthesis control unit (39) It is characterized by...

2. Multiple visual data processing and feedback virtual makeup simulation according to Claim 1. The system (1) has the feature of obtaining different viewing and lighting conditions. Storing visual data within a Multiple visual dataset (3), the visual data in question The processing of data by the Image Processing Unit (4) and the visual data and words 30 The subject is Data, at least one of which is intermediate or final data obtained from visual data. It is characterized by being stored in the storage unit (5).

3. Multiple visual data processing and feedback virtual makeup simulation according to Claim 1. The system (1) has the feature that the imaging condition analysis unit (9) is different. 47 Analysis of viewing conditions related to visual data, Lighting The lighting in the said visual data by the feature determination unit (10) Determination of characteristics, Exposure and shading determination unit (11) exposure and shadowing effects on the face or skin areas by determination and at least one of the data points in question is Regional Image 5 usability determination unit (12) by face or different visual data in different visual data Used in regionally determining the usability of skin areas. It is characterized by...

4. Multiple visual data processing and feedback virtual makeup simulation according to Claim 3. The system (1) has the following feature: Regional feature reliability determination unit (18) 10 by the Inter-Image Feature Variability Determination Unit (17) Regional image with feature variability between specified images. regional image determined by the availability determination unit (12) joint assessment of usability and determination of lighting features. Exposure and shadowing with lighting characteristics determined by unit (10) 15 the exposure and shading effects determined by the determination unit (11) Regional characteristic reliability by including at least one in the assessment It is characterized by the creation of value (19).

5. Multiple visual data processing and feedback virtual makeup simulation according to Claim 4. The system (1) has the following feature; Regional feature reliability value (19) value, 20 inter-image feature variability of the corresponding face or skin region increase, regional image availability decrease, lighting conditions negatively affecting the determination of the relevant surface characteristic or exposure and In at least one of the situations where shading effects are increased, the relevant surface property 25 performed by the reliability weighted feature combination unit (20) will reduce its contribution to the feature merging process; features between images reduced variability or increased regional image availability in at least one of these situations, it should be created in a way that will increase the contribution in question. It is characterized by...

6. Multiple visual data processing and feedback virtual makeup simulation according to Claim 4 30 The system (1) has the feature; Surface feature extraction unit (13) is matched by. Regional color feature data from face or skin areas (14), Regional tone and Brightness feature data (15) and Regional contrast and surface appearance feature data (16) creation and Reliability weighted feature combination unit (20) 48 by a combined regional of each of the regional characteristics in question The contribution of the surface feature to the relevant Regional feature reliability value (19) is based on It is characterized by being determined separately by taking into account the data.

7. Multiple visual data processing and feedback virtual makeup simulation according to Claim 6. The system (1) has the feature of; Stable surface feature creation unit (21) by 5 reliability differs primarily from combined regional surface features in visual terms. regional surface characteristics that show continuity among the data the creation and relevant regional surface characteristics of the face or skin User-specific surface feature data structure (22) associated with regions It is characterized by its representation within it. 10 8. Multiple visual data processing and feedback virtual makeup simulation according to Claim 1. The system (1) has the feature of; Visual context analysis unit (23) by visual Analysis of the context, clothing by the Clothing Zone Determination Unit (24) Determining the region, clothing color distribution from the clothing region in question. Creating the data (25), Environment visual feature data relating to the viewing environment 15 (26) determination and Clothing color distribution data (25), Environment visual feature data (26) and User-specific surface feature data structure (22) are processed together. It is characterized by the creation of a context feature data structure (27).

9. Multiple visual data processing and feedback virtual makeup simulation according to Claim 8. The system (1) has the following feature: Regional image synthesis parameter determination unit 20 (28) User-specific surface feature data structure (22) with Unified context feature data structure (27) evaluated together for different face or skin areas characteristics of color, tone, brightness, density, transparency, regional distribution and transitions Regional image synthesis parameters (29) which determine at least one of them creation and User-specific surface feature data structure (22) or Combined 25 A regional change within the context feature data structure (27) affects the relevant face or At least one of the regional image synthesis parameters (29) of the skin region It is characterized by being transferred to someone's value.

10. Multiple visual data processing and feedback virtual makeup simulation according to Claim 9. The system (1) has the feature of; Light 30 by Target lighting modeling unit (30). intensity, direction of incidence, and color characteristic of the light, representing at least one of these. The creation of target illumination data (31) and the said target illumination data (31) value is represented in the User-specific surface feature data structure (22). different faces or skin tones, taking into account the regional surface characteristics observed 49 Regionally, in a way that will produce different image synthesis results in different regions. Virtual image synthesis parameters (29) are regionally correlated with the Virtual The image is synthesized by the makeup image synthesis unit (32). It is characterized by its use as an input in the synthesis process.

11. Multiple visual data processing and feedback virtual makeup simulation according to claim 10 5 The system (1) has the feature of; by the Regional visual consistency control unit (34). Synthesized virtual makeup image (33) of the relevant face or skin User-specific surface properties with redefined image characteristics from the region. surface features associated with the relevant region within the data structure (22), Regional image synthesis parameters (29), Target illumination data (31) and 10 at least one of the image features relating to adjacent facial or skin areas the difference between the regional reference image feature determined using identification and determination of the difference in question regarding the relevant facial or skin area, visual feature Regional visual discrepancy value (35) representing the type and amount of difference It is characterized by its creation as follows: 15 12. Multiple visual data processing and feedback virtual makeup simulation according to Claim 11. The system (1) has the feature of; Incompatibility classification and parameter matching unit. Regional visual incongruity value (35) represented by (36) the discrepancy in color, tone, brightness, intensity, transparency, regional distribution or Classification based on at least one of the transition characteristics and the 20 classified items modifying the incompatibility in order to reduce the incompatibility in question At least one of the required Regional image synthesis parameters (29) Parameter matching; Parameter correction data due to mismatch. The regional visual incompatibility value (35) is mentioned by the creation unit (37). The subject is matched using the relevant facial or skin area and the relevant Regional 25. parameter-specific correction within image synthesis parameters (29) It is characterized by the creation of data.

13. Multiple visual data processing and feedback virtual makeup according to Claim 12. simulation system (1) and its feature is; Feedback image synthesis Parameter update unit (38) Parameter 30 due to incompatibility to the correction data generated by the correction data generation unit (37) depending on the relevant Regional image synthesis parameters (29) at least updating one of the updated Regional image synthesis parameters (29) A new virtual makeup image synthesis unit (32) is created by using 50 Creating a synthesized virtual makeup image (33) is the subject of this new Regional visual consistency on synthesized virtual makeup image (33) Regional visual consistency audit by audit unit (34) by creating the current Regional visual incompatibility value (35) and Consistency condition assessment and resynthesis control unit (39) 5 by the current Regional visual mismatch value (35) data Incompatibility classification if it fails to meet the defined consistency requirement. parameter matching, parameter correction, parameter update and reset until the consistency condition of the synthesis processes is met or a defined process It is characterized by being repeated until the control limit is reached. 10 14. Data obtained for the same user under different viewing and lighting conditions. virtual processing of multiple visual data using computer applications a virtual makeup simulation for creating a makeup look The method and its characteristic is; different display and lighting for the same user. Obtaining multiple visual data under these conditions; the visual in question is 15 Identifying facial regions within the data and the identified facial regions segmentation of skin regions within; within different visual data Matching corresponding facial and skin regions; matched facial and skin Extraction of regional surface features from regions; corresponding face or skin 20 differences between surface features extracted from different regions and visual data. Determining the variability of facial or skin regions in different visual data; related lighting characteristics and at least one of the exposure and shadowing effects Determining regional image availability through evaluation; word The subject is regional imagery with inter-image feature variability. Surface properties derived by jointly evaluating their usability 25 Creating regional feature reliability values ​​for; different visual data surface features extracted from corresponding facial or skin areas the contribution of the combined regional surface feature to the regional feature in question Weighting based on reliability values; reliability weighted User-specific surface feature data from combined regional surface features 30 Creating the structure; color distribution and visualization for clothing areas. Identifying at least one of the visual characteristics of the environment and the visual in question contextual features along with a user-specific surface feature data structure. Creating a unified context feature data structure by processing; providing to the user 51 the combination of a surface feature data structure and a context feature data structure. regional images for different facial or skin areas by evaluation Establishing synthesis parameters; representing the targeted lighting condition. Generating target lighting data; user-specific surface feature data structure, regional image synthesis parameters and target illumination data 5 Synthesizing a virtual makeup image using; synthesized virtual makeup Re-identified face or skin areas within the image with its features, the user-specific surface feature data structure, and the relevant regional image. synthesis parameters, target illumination data and adjacent face or skin 10 determined using at least one of the image features relating to the regions Evaluation of the difference between regional reference image characteristics the facial or skin area in question, the type of visual feature, and the degree of difference. the creation of a regional visual dissonance value representing; the incongruity represented by the regional visual incongruity value 15 that need to be changed in order to reclassify and reduce inconsistencies matching with at least one of the regional image synthesis parameters; word The relevant face is determined based on the subject matching and regional visual incompatibility value. or specific to the skin region and the relevant regional image synthesis parameter. Generating parameter correction data due to discrepancies; generated Depending on the parameter correction data due to the discrepancy, the relevant regional 20 Updating at least one of the image synthesis parameters; updated virtual makeup image using regional image synthesis parameters resynthesis; on the resynthesized virtual makeup image The regional visual consistency check will be re-conducted and the current regional Determining the visual incompatibility value; the relevant current regional visual 25 whether the discrepancy value satisfies the defined consistency condition evaluation and if the aforementioned consistency condition is not met regional visual consistency check, regional visual inconsistency value identification, mismatch classification and parameter matching, mismatch Generating source parameter correction data, regional image synthesis 30 updating the parameters and defining the resynthesis processes until the consistency condition is met or a defined transaction control limit is reached. It is characterized by including steps that are repeated until the desired result is achieved. is being done. 52 15. According to claim 14, it is a virtual makeup simulation method characterized by its different visual features. Analysis of imaging conditions relating to facial or skin areas in the data this involves considering the lighting characteristics and exposure in those areas. Determining the effects of shading and identifying at least one of the data points that is different. 5 of the regional image availability of face or skin areas in visual data It is characterized by its use in determination.

16. According to claim 15, it is a virtual makeup simulation method, characterized by its corresponding face. or surface features extracted from skin areas, compared with different visual data. by determining the variability and the feature variability between the images in question. The regional image availability specified under Claim 15 is 10 together. evaluation, as well as lighting characteristics, exposure and shadowing by including at least one of its effects in the assessment in question. regional representation of the usability of surface feature in different visual data The feature is characterized by the creation of a reliability value.

17. According to claim 16, it is a virtual makeup simulation method and its characteristic is; regional characteristic 15 the reliability value between images relating to the corresponding face or skin area. Increased feature variability, decreased regional image availability, lighting conditions negatively affect the determination of the relevant surface characteristic. or at least one of the following situations is related to increased exposure and shadowing effects. the contribution of surface feature to reliability-weighted feature combining process 20 will reduce; reduction of feature variability between images or regional image in at least one of the situations where its availability increases, the contribution in question It is characterized by being determined in a way that will increase it.

18. According to claim 16, it is a virtual makeup simulation method characterized by; matched faces. or regional color characteristics, regional tone and brightness characteristics from skin areas 25 Extraction of regional contrast and surface appearance features and different visuals the data were extracted from the corresponding facial or skin areas the contribution of each of the features to the combined regional surface feature by determining them separately based on regional characteristic reliability values. It is characterized by 30 19. According to claim 18, it is a virtual makeup simulation method characterized by its reliability. visual data differing from regional surface features that are predominantly combined the creation of regional surface characteristics that show continuity among them and the regional surface characteristics in question with the relevant facial or skin areas 53 By associating them, the user-specific surface feature data structure is represented within the data. It is characterized by its...

20. According to claim 19, it is a virtual makeup simulation method characterized by its visual context. analysis, identification of the clothing area, the clothing area in question Determining the color distribution of clothing in relation to the viewing environment, environment 5 determining the visual characteristics and the color distribution of the clothing in question and the environment processing visual features together with a user-specific surface feature data structure. Characterized by the creation of a combined context feature data structure. is being done.

21. According to claim 20, it is a virtual makeup simulation method with the feature of being user-specific. 10 surface feature data structure and combined context feature data structure together color, tone, and brightness for different facial or skin areas are determined by evaluation. density, transparency, regional distribution, and transition properties, at least one of them and the creation of regional image synthesis parameters that determine user-specific surface feature data structure or combined context feature data structure 15 a regional change within it, regionally related to the relevant facial or skin area by transferring the value of at least one of the image synthesis parameters It is characterized by...

22. According to claim 21, it is a virtual makeup simulation method, the characteristic of which is; the targeted light intensity, direction of incidence and color of light related to lighting conditions 20 Generating target lighting data that represents at least one of its characteristics and the target lighting data in question, user-specific surface feature data taking into account the regional surface characteristics represented within its structure Different facial or skin areas produce different image synthesis results. virtual 25 with regional image synthesis parameters in a way that will bring It is characterized by its use as input in the synthesis of makeup looks. is being done.

23. According to claim 22, it is a virtual makeup simulation method, characterized by its synthesized properties. virtual makeup image redefined from the relevant face or skin area The relevant 30 within the user-specific surface feature data structure with image feature. surface features associated with the region, relevant regional image synthesis parameters, target illumination data and adjacent facial or skin areas A regional reference image determined using at least one of the image features. determining the difference between the features and the occurrence of that difference 54 the facial or skin area it comes from, the type of visual feature it is associated with, and the differences by creating a regional visual discrepancy value that represents the amount It is characterized by...

24. According to claim 23, it is a virtual makeup simulation method, characterized by its regional visualization. The mismatch, represented by the mismatch value, is a variation in color, tone, brightness, and intensity. in terms of at least one of the characteristics of transparency, regional distribution or transition classification of the classified nonconformity; the nonconformity in question regional image synthesis that needs to be modified in order to reduce matching with at least one of its parameters; the face where the mismatch occurs or regional image matched with skin area, type and amount of discrepancy 10 based on the synthesis parameter, specific to the region and parameter in question. Generating parameter correction data due to discrepancies; generated relevant regional parameter correction data due to discrepancy Updating at least one of the image synthesis parameters; updated 15 virtual makeup images using regional image synthesis parameters resynthesis; on the resynthesized virtual makeup image The regional visual consistency check will be re-conducted and the current regional Determining the visual incompatibility value; the current regional visual in question. whether the discrepancy value satisfies the defined consistency condition evaluation and if the aforementioned consistency condition is not met, 20 regional visual consistency check, regional visual inconsistency value identification, mismatch classification and parameter matching, mismatch Generating source parameter correction data, relevant regional image Updating and resynthesizing at least one of the synthesis parameters until the defined consistency condition of the transactions is met or a defined transaction 25 It is characterized by being repeated until the control limit is reached.

Images