Intelligent beauty method based on face partition

Abstract

The invention discloses an intelligent beauty method based on face partition, and relates to the technical field of face intelligent beauty, an image acquisition module is used for acquiring an image of a user face partition, an image processing module is used for calculating and outputting an adjusted skin brightness value LF, and a skin simulation module is used for performing simulation deduction on the state of the optimized skin, so that the skin brightness value LF is obtained. The image processing module is used for respectively calculating and outputting optimized skin smoothness PF and skin color uniform value JF, extracting a skin color uniform threshold value JFHZ corresponding to the age group of the patient from the database, comparing the skin color uniform threshold value JFHZ with the skin color uniform value JF, and outputting a beautifying method by the interaction module. By means of accurate face partition processing, a scientific parameter adjustment mechanism, an efficient iteration processing flow and remarkable skin color uniformity improvement, the method not only improves the processing precision and efficiency of the intelligent beauty treatment method, but also provides a more personalized and accurate beauty treatment scheme for a user.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent facial beauty, and in particular to an intelligent beauty method based on facial zoning. Background Art

[0002] With the continuous development of image processing technology, the application of smart beauty methods in the beauty industry is becoming more and more extensive. Among them, facial zoning is an important concept in the field of cosmetic medicine and skin care. According to the theory of facial aesthetic zoning, the human face can be divided into multiple parts, such as forehead, eyebrows, nose, lips, chin, nasolabial groove, eye sockets, cheeks, and temporal parts. The skin characteristics, aging manifestations and required care methods of these parts are different. Therefore, the smart beauty method based on facial zoning can provide precise care for different parts to meet personalized beauty needs.

[0003] However, some existing intelligent beauty methods often adopt a global processing strategy, without personalized processing for different facial areas, resulting in less precise processing results. In traditional beauty methods, parameter adjustment often relies on manual experience, and thus lacks scientific basis and automated adjustment mechanism. When performing multiple iterative processes, existing methods often find it difficult to quickly converge to the optimal solution, resulting in low iteration efficiency. In addition, in terms of processing skin color uniformity, existing methods often find it difficult to accurately identify and process facial spots and pores, resulting in poor results in improving skin color uniformity. Summary of the invention

[0004] The purpose of the present invention is to provide an intelligent beauty method based on facial partitioning, which solves the problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solution, and the specific implementation steps are as follows: Step 1: Use the image acquisition module to collect images of the user's facial partitions, transmit them to the image processing module, and store them in the database; Step 2: Using the image processing module, calculate and output the adjusted skin brightness value LF; Step 3: Based on the adjusted skin brightness value LF and the custom brightness boost value TL, the skin simulation module is used to simulate the optimized skin state and summarize the noise pixel number XS. noise ; Step 4: XS based on the number of noise pixels noise and the adjusted skin brightness value LF, and using the image processing module, respectively calculate and output the optimized skin smoothness PF and skin color uniformity value JF; Step 5: Extract the skin color uniformity threshold JF corresponding to the patient's age group from the database HZ, and compare it with the skin color uniformity value JF, and the interactive module outputs the beauty method; The image processing module includes a unit for improving the brightness of the skin in facial zones, a unit for optimizing the smoothness of the skin in facial zones, and a unit for improving the uniformity of the skin in facial zones.

[0006] Preferably, the equipment used by the image acquisition module includes a high-resolution camera and professional photography equipment; The equipment used in the image processing module includes a computer and image processing software; The equipment used in the skin simulation module includes a skin simulation system; The devices used by the interaction module include a display device.

[0007] Preferably, the calculation formula for improving the skin brightness unit of the facial partition is as follows: LF=SQRT(R avg +G avg +B avg )×(1+TL / 100)-Y / 2; Y=YM / FM; in: LF is the adjusted skin brightness value; SQRT is the square root symbol; R avg The average redness of the skin; G avg Average skin glossiness; B avg Average skin transparency; TL is the custom brightness boost value. TL reflects the degree to which the user-defined brightness needs to be boosted. Its value range is {0-100}; Y is the skin shading degree; YM is the shaded area, and FM is the total area of ​​the facial partitions.

[0008] Preferably, the average redness of the skin R avg The calculation formula is as follows: R avg =(R1+R2+R3+......+R N ) / N; N is the total number of pixels, which reflects the total number of pixels for skin detection in the facial area; R1 is the skin redness of the first pixel, R2 is the skin redness of the second pixel, R3 is the skin redness of the third pixel, R N is the skin redness of the Nth pixel; The average glossiness of the skin G avg The calculation formula is as follows: G avg =(G1+G2+G3+......+G N ) / N; G1 is the skin glossiness of the first pixel, G2 is the skin glossiness of the second pixel, G3 is the skin glossiness of the third pixel, G N is the skin glossiness of the Nth pixel; The average transparency of the skin B avg The calculation formula is as follows: B avg =(B1+B2+B3+......+B N ) / N; B1 is the skin transparency of the first pixel, B2 is the skin transparency of the second pixel, B3 is the skin transparency of the third pixel, N The Nth pixel skin transparency.

[0009] Preferably, based on the custom brightness enhancement value TL, after simulating brightness enhancement on the face area, the average redness of the skin R avg , Average skin glossiness G avg and skin average transparency B avg will change, and then calculate the average redness of the skin after optimization R avg ’ , Average skin glossiness after optimization G avg ’ and average skin transparency after optimization B avg ’ , and the average redness of the skin R avg , Average skin glossiness G avg and skin average transparency B avg The number of noise pixels with large differences from the average pixel in the facial area is calculated as follows: According to HZ R =SQRT(R avg -R avg ’ ) 2 , HZ G =SQRT(G avg -G avg ’ ) 2 and HZ B =SQRT(B avg -B avg ’ ) 2 ) to calculate the redness difference threshold HZ R , gloss difference threshold HZ G and transparency difference threshold HZB And the skin redness R of the i-th pixel before and after optimization within the total number of pixels N is calculated. i , the skin glossiness G of the i-th pixel i and the skin transparency B of the i-th pixel i , compare one by one, and exceed the redness difference threshold HZ R , gloss difference threshold HZ G and transparency difference threshold HZ B The pixels with large differences are counted as the number of noise pixels with large differences, that is, the number of noise pixels after aggregation XS noise .

[0010] Preferably, the calculation formula for optimizing the facial partition skin smoothness unit is as follows: PF=LF×(1-XS noise / N)+SQRT(LF×R std ) / 10-XS pore / (LF×2); ; in: PF is the skin smoothness after optimization; XS noise is the number of noise pixels; R std is the standard deviation of skin redness; XS pore is the number of pore pixels.

[0011] Preferably, the calculation formula for improving the skin uniformity unit of the facial zone is as follows: JF=PF×[1-|R avg -G avg | / (R avg +G avg +1)]+SQRT(LF / (XS pore +1)-PF×(SB / XS pore ); in: JF is the skin color uniformity value; SB is the spot area.

[0012] Preferably, the adjustment steps of the smart beauty method based on the skin color uniformity value JF are as follows: S1. Determine a custom brightness enhancement value TL according to user needs; S2, calculate the adjusted skin brightness value LF, the optimized skin smoothness PF and the skin color uniformity value JF in sequence; S3, the skin color uniformity value JF and the skin color uniformity threshold JF HZ Make a comparison If the skin color uniformity value JF is higher than the skin color uniformity threshold JF HZ , it reflects that the skin color of the facial area is too uniform and easy to lose the natural feeling. When the pores and spots are reduced and optimized, the spot area SB and the number of pore pixels XS are increased. pore ; If the skin color uniformity value JF is lower than the skin color uniformity threshold JF HZ , it reflects that the texture of the facial partition changes dramatically. When the area of ​​pores and spots is increased and optimized to reduce the spot area SB and the number of pore pixels XS pore ; If the skin color uniformity value JF is equal to the skin color uniformity threshold JF HZ , it reflects that the facial zoning beauty method achieves the effect when the current beauty method is performed.

[0013] Preferably, the database stores skin color uniformity values ​​JF and skin color uniformity thresholds JF for different age groups. HZ , and regularly adjust the skin color uniformity threshold JF HZ Update, where the skin color uniformity threshold JF HZ Specifically, it is the average value of skin color uniformity value JF of different age groups, and when adjusting and comparing beauty methods, the skin color uniformity threshold JF of the corresponding age group is extracted HZ .

[0014] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention improves the processing accuracy and effect by accurately segmenting the facial area of ​​the image and performing personalized beauty processing on each sub-area. This is due to the detailed division and targeted processing of different facial areas in the image processing module, which includes a facial zone skin brightness enhancement unit, a facial zone skin smoothness optimization unit, and a facial zone skin uniformity enhancement unit.

[0015] Second, the present invention realizes the automatic adjustment and scientific calculation of parameters by constructing three interrelated algorithm formulas, and in each iteration, automatically adjusts the spot area SB and the number of pore pixels XS according to the result of the skin color uniformity value JF pore , thus achieving rapid convergence of parameters and acquisition of the optimal solution.

[0016] 3. The present invention can quickly converge to the optimal solution through continuous iteration and adjustment of parameters, thereby improving the efficiency of iterative processing. At the same time, by introducing a cyclic influence mechanism, the cyclic iteration and optimization of the three formulas of the adjusted skin brightness value LF, the optimized skin smoothness PF and the skin color uniformity value JF are realized, and the processing effect is further improved.

[0017] Fourth, the present invention significantly improves the uniformity of skin color by accurately identifying and processing facial spots and pores. In the unit for improving the uniformity of facial skin, the spot area SB and the number of pore pixels XS are introduced. pore Parameters are used to effectively deal with uneven skin tone. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A flowchart of the method of the intelligent beauty method based on facial partitioning; Figure 2 It is a structural schematic diagram of the image processing module of the present invention; Figure 3 is the number of noise pixels XS in the present invention noise A summary diagram of Figure 4 Schematic diagram of adjustment feedback of skin color uniformity value JF of the present invention. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] This intelligent beauty method based on facial zoning is different from existing beauty methods. Existing beauty methods have problems such as insufficient processing accuracy, difficulty in parameter adjustment, low iteration efficiency, and poor skin color uniformity. This algorithm unit effectively solves the problems and shortcomings of the existing technology through precise facial zoning processing, scientific parameter adjustment mechanism, efficient iterative processing flow and significant improvement in skin color uniformity. This method not only improves the processing accuracy and efficiency of the intelligent beauty method, but also provides users with a more personalized and accurate beauty plan.

[0021] For example, see Figures 1 to 4 This implementation provides a smart beauty method based on facial partitioning, and the specific implementation steps are as follows: Step 1: Use the image acquisition module to collect images of the user's facial partitions, transmit them to the image processing module, and store them in the database; Step 2: Using the image processing module, calculate and output the adjusted skin brightness value LF; Step 3: Based on the adjusted skin brightness value LF and the custom brightness boost value TL, the skin simulation module is used to simulate the optimized skin state and summarize the noise pixel number XS. noise ; Step 4: XS based on the number of noise pixels noise and the adjusted skin brightness value LF, and using the image processing module, respectively calculate and output the optimized skin smoothness PF and skin color uniformity value JF; Step 5: Extract the skin color uniformity threshold JF corresponding to the patient's age group from the database HZ , and compare it with the skin color uniformity value JF, and the interactive module outputs the beauty method; The image processing module includes a unit for improving the brightness of the skin in facial zones, a unit for optimizing the smoothness of the skin in facial zones, and a unit for improving the uniformity of the skin in facial zones; The equipment used in the image acquisition module includes high-resolution cameras and professional photography equipment; The equipment used in the image processing module includes computers and image processing software; The equipment used in the skin simulation module includes a skin simulation system; The devices used by the interactive module include display devices.

[0022] In this embodiment, the system forms the core of intelligent beauty effect optimization through the mutual cooperation of three algorithm units and the combination of the three operation results of LF, PF and JF. Specifically, LF is the adjusted skin brightness value, which aims to enhance the overall brightness of the skin while taking into account the influence of the shadow area to achieve a more uniform and bright skin tone effect. PF is the optimized skin smoothness, which aims to reduce blemishes and texture changes on the skin to make the skin look smoother and more delicate. JF is the skin tone uniformity value, which aims to reduce the influence of spots and pores on the skin tone to make the skin tone more uniform and natural. The calculation result of JF can also affect the calculation feedback to LF and PF, so that the three algorithms of the system can achieve comprehensive optimization of skin brightness, smoothness and skin tone uniformity through continuous iteration and adjustment of parameters, and provide users with a more personalized and efficient beauty experience.

[0023] See also Figures 1 to 4 , the calculation formula for improving the skin brightness unit of the facial zone is as follows: LF=SQRT(R avg +G avg +B avg )×(1+TL / 100)-Y / 2; Y=YM / FM; in: LF is the adjusted skin brightness value; SQRT is the square root symbol; R avg The average redness of the skin; G avg Average skin glossiness; B avg Average skin transparency; TL is the custom brightness boost value. TL reflects the degree to which the user-defined brightness needs to be boosted. Its value range is {0-100}; Y is the skin shading degree; YM is the shaded area, FM is the total area of ​​the facial partitions; Average skin redness R avg The calculation formula is as follows: R avg =(R1+R2+R3+......+R N ) / N; N is the total number of pixels, which reflects the total number of pixels for skin detection in the facial area; R1 is the skin redness of the first pixel, R2 is the skin redness of the second pixel, R3 is the skin redness of the third pixel, R N is the skin redness of the Nth pixel; Average skin glossiness G avg The calculation formula is as follows: G avg =(G1+G2+G3+......+G N ) / N; G1 is the skin glossiness of the first pixel, G2 is the skin glossiness of the second pixel, G3 is the skin glossiness of the third pixel, G N is the skin glossiness of the Nth pixel; Average skin transparency B avg The calculation formula is as follows: B avg =(B1+B2+B3+......+B N ) / N; B1 is the skin transparency of the first pixel, B2 is the skin transparency of the second pixel, B3 is the skin transparency of the third pixel, N The Nth pixel skin transparency.

[0024] In this embodiment: First, in this algorithm unit, "SQRT(R avg +G avg +B avg The calculation part of the ")" calculates the root result of the sum of the average values ​​of the red, green and blue channels of the face area of ​​the image, and the root result of this average value reflects the overall brightness level of the face area, and serves as the basis of the calculation formula for improving the skin brightness unit of the face partition. It determines the baseline brightness before brightness adjustment, and by combining it with the user-defined custom brightness boost value TL, the adjusted skin brightness value LF can be calculated; The "(1+TL / 100)" calculation part is used to adjust the baseline brightness according to the user-defined custom brightness boost value TL. By multiplying the "(1+TL / 100)" calculation part, a linear increase in brightness can be achieved. This calculation part determines the magnitude of the brightness adjustment, that is, the percentage by which the user wants the skin brightness to increase; The "Y / 2" calculation part takes into account the effect of the skin shadow Y in the facial area of ​​the image on the brightness adjustment. By subtracting the "Y / 2" calculation part, the effect of the shadow area on the overall brightness can be simulated, making the adjusted brightness more natural. This calculation part is used to adjust the brightness to take into account the existence of facial shadows and avoid over-brightening the shadow area; This algorithm unit allows users to fine-tune skin brightness according to their aesthetic preferences through the custom brightness boost value TL. The setting of this parameter makes the beauty treatment more flexible and can meet the different brightness requirements of different users. Specifically, for users who want brighter skin, the value of the custom brightness boost value TL can be appropriately increased, while for users who prefer natural skin color, the custom brightness boost value TL can be kept at a lower level; In addition, the introduction of skin shadow Y enables the facial zone skin brightness enhancement unit to more accurately locate and process the shadow areas on the face. By increasing the brightness of these areas, the overall brightness uniformity of the face can be significantly improved, and the shadow effect caused by uneven light can be reduced. This treatment not only improves the visual effect of the skin, but also makes the facial contour clearer and enhances the three-dimensional effect. Calculation of the average values ​​of the red, green and blue channels ensures color balance. During the processing, increasing the skin brightness unit of the facial zone ensures that the brightness changes between the three channels remain coordinated, avoiding color distortion caused by adjusting a single channel. This color balance processing makes the skin look more natural and healthy.

[0025] See also Figures 1 to 4 , the calculation formula for optimizing the skin smoothness unit of facial partition is as follows: PF=LF×(1-XS noise / N)+SQRT(LF×R std ) / 10-XS pore / (LF×2); ; in: PF is the skin smoothness after optimization; XS noise is the number of noise pixels; R std is the standard deviation of skin redness; XS pore is the number of pore pixels.

[0026] In this embodiment, first, "LF×(1-XS noise / N)” calculation part calculates the number of noise pixels XS noise The ratio relative to the total number of pixels N is used to adjust the adjusted skin brightness value LF. By reducing the influence of noise pixels on the adjusted skin brightness value LF, the skin smoothness can be optimized. It determines the influence of noise on the smoothness optimization, and by reducing the interference of noise, the skin looks smoother. “SQRT(LF×R std ) / 10” calculation part takes into account the standard deviation R of the skin redness of the red channel in the facial area of ​​the image std , that is, the degree of texture change, is multiplied by the adjusted skin brightness value LF and divided by the sum 10, where 10 is a weight coefficient used to adjust the influence of texture change on smoothness, thereby simulating the influence of texture change on smoothness. It is used to consider texture changes in smoothness optimization to avoid excessive smoothing leading to loss of skin details; “XS pore / (LF×2)” Even if the number of pore pixels is partially taken into account, pore Effect on smoothness, by number of pore pixels XS pore Dividing the adjusted skin brightness value LF by 2 can reduce the negative impact of pores on smoothness, where 2 is a weight coefficient used to adjust the degree of influence of pores on smoothness. This calculation part is used to reduce the interference of pores in smoothness optimization and make the skin look more delicate. This algorithm unit passes the noise pixel number XS noise The ratio of the total number of pixels N reflects the density of noise in the image. The optimized facial partition skin smoothness unit can accurately identify and suppress the noise in the image by calculating this ratio. This processing not only reduces the blemishes on the skin, but also makes the skin look smoother and more delicate. The standard deviation of skin redness R std Reflects the texture changes of the skin. Optimizing the skin smoothness of facial partitions can improve the texture of the skin by adjusting this parameter, making it look more natural and uniform. This texture optimization process makes the skin more delicate and elastic visually, and the number of pore pixels is XS pore Consideration enables the optimization of facial zone skin smoothness unit to perform refinement processing on pores. By reducing the number of pixels of pores, the fineness of the skin can be significantly improved, making it look firmer and smoother. This treatment is undoubtedly a great blessing for users who pursue delicate skin.

[0027] See also Figures 1 to 4, the calculation formula for improving the skin uniformity unit of facial zones is as follows: JF=PF×[1-|R avg -G avg | / (R avg +G avg +1)]+SQRT(LF / (XS pore +1)-PF×(SB / XS pore ); in: JF is the skin color uniformity value; SB is the spot area.

[0028] In this embodiment, the algorithm unit first “PF×[1-|R avg -G avg | / (R avg +G avg +1)]” The calculation part considers the ratio of the difference between the average values ​​of the red and green channels to the sum of the two, which is used to simulate the degree of uneven skin color. By adjusting this ratio, the uniformity of the skin color can be optimized. This calculation part determines the impact of the uneven skin color on the uniformity improvement, and makes the skin color more uniform by reducing the negative impact of the difference between the red and green channels on the skin color uniformity value JF; “SQRT(LF / (XS pore +1)” calculation part takes into account the adjusted skin brightness value LF relative to the number of pore pixels XS pore By calculating this ratio, the uniformity of skin color can be further adjusted, especially for the pore area. This calculation part is used to consider the influence of pores in uniformity improvement, and make the skin look more uniform by adjusting the brightness difference of the pore area; "PF×(SB / XS pore The calculation part takes into account the effect of the spot area SB on the uniformity of skin color, by subtracting the optimized skin smoothness PF multiplied by the spot area SB and the number of pore pixels XS pore The ratio can reduce the negative impact of spots on skin tone uniformity. It is used to reduce the interference of spots during uniformity improvement, making the skin tone look more even and consistent; In this algorithm unit, the adjustment of the red channel average and the green channel average can directly affect the hue and saturation of the skin color. The facial partition skin uniformity improvement unit can significantly improve the uniformity and naturalness of the skin color by calculating these two parameters and making appropriate adjustments to them. This skin color correction process makes the skin look healthier and more radiant. The consideration of the spot area SB enables the facial zone skin uniformity enhancement unit to treat spots. By calculating the spot area SB and adjusting it appropriately, the effect of spots on skin color can be significantly reduced and the uniformity of skin color can be improved. This treatment is undoubtedly an effective solution for users with spot problems. In addition, the unit for improving the uniformity of skin in facial zones not only takes into account the processing of skin brightness, smoothness and spot reduction, but also achieves comprehensive optimization of skin color by combining the results of these parameters. This comprehensive optimization processing makes the skin more visually uniform and natural, improving the overall beauty effect.

[0029] In summary, the beneficial effects of the facial zone skin brightness enhancement unit, the facial zone skin smoothness optimization unit and the facial zone skin uniformity enhancement unit are closely combined with the parameters and their functions. The setting and adjustment of these parameters make the beauty treatment more precise and flexible, and can meet the different needs of different users for skin beauty.

[0030] See also Figures 1 to 4 The adjustment steps of the smart beauty method based on the skin color uniformity value JF are as follows: S1. Determine a custom brightness enhancement value TL according to user needs; S2, calculate the adjusted skin brightness value LF, the optimized skin smoothness PF and the skin color uniformity value JF in sequence; S3, the skin color uniformity value JF and the skin color uniformity threshold JF HZ Make a comparison If the skin color uniformity value JF is higher than the skin color uniformity threshold JF HZ , it reflects that the skin color of the facial area is too uniform and easy to lose the natural feeling. When the pores and spots are reduced and optimized, the spot area SB and the number of pore pixels XS are increased. pore ; If the skin color uniformity value JF is lower than the skin color uniformity threshold JF HZ , it reflects that the texture of the facial partition changes dramatically. When the area of ​​pores and spots is increased and optimized to reduce the spot area SB and the number of pore pixels XS pore ; If the skin color uniformity value JF is equal to the skin color uniformity threshold JF HZ , it reflects that the facial zoning beauty method achieves the effect when the current beauty method is performed.

[0031] In this embodiment, the algorithm unit can dynamically adjust the color spot area SB and the number of pore pixels XS according to the skin color uniformity value J calculated by the facial partition skin uniformity enhancement unit. pore, making skin brightness adjustment more accurate. Through multiple iterations, the results of improving the skin brightness unit of the facial zone and improving the skin uniformity unit of the facial zone will gradually converge to achieve a satisfactory level of skin color uniformity, avoiding the problem of over-adjustment or under-adjustment. In addition, the circular influence mechanism enables the skin brightness unit of the facial zone and the skin uniformity unit of the facial zone to cooperate with each other to jointly achieve the overall optimization of skin brightness, smoothness and skin color uniformity, thereby improving the effect of the smart beauty method and user experience. In addition, when the custom brightness boost value TL is defined too high, resulting in the skin color uniformity value JF showing uneven facial partitions, the parameters in the formula can be dynamically adjusted by adjusting the shadow area hand-grip ratio, optimizing noise processing, adjusting the number of pore pixels and the color spot area SB strategy, thereby optimizing skin color uniformity. This adjustment is based on real-time feedback, and can then be accurately processed for specific problems. Through the loop feedback mechanism, the algorithm can continuously receive processed image data and adjust parameters according to the display results of the skin color uniformity value JF. This continuous optimization process can gradually approach the ideal skin color uniformity state and improve the accuracy and reliability of the algorithm. During the adjustment process, the algorithm will take multiple factors into consideration to avoid excessive processing that will make the skin tone too uniform and lose its natural feel. By fine-tuning the parameters, the algorithm can improve the uniformity of the skin tone while maintaining the natural texture and details of the skin. If the uniformity value JF is too high, the algorithm will try to increase the spot area SB and the number of pore pixels XS. pore , to restore natural unevenness in skin tone. This adjustment helps keep the skin looking lively and realistic. The parameter adjustment strategy in the algorithm can adapt to skin conditions of different age groups, different skin types and different lighting conditions. By flexibly adjusting parameters, the algorithm can accurately process various skin conditions, improve the adaptability of the algorithm, and thus maintain stable performance in the face of various complex situations. This robustness helps the algorithm cope with various challenges and uncertainties in practical applications. Through the loop feedback mechanism, the algorithm can customize personalized beauty plans based on the user's personal preferences and skin conditions. This personalized plan can more accurately meet the needs of users and improve user satisfaction and loyalty. The loop feedback mechanism enables the algorithm to receive user feedback in real time and make adjustments. This interactivity helps to enhance the user's sense of participation and trust, and improve the overall level of user experience. In summary, combined with improving the facial zone skin brightness unit, optimizing the facial zone skin smoothness unit, and improving the facial zone skin uniformity unit and its parameters, the circular feedback mechanism formed by the above adjustment strategy has brought significant beneficial effects in improving the accuracy of the skin color uniformity value JF, maintaining the natural feel of the skin, improving the adaptability and robustness of the algorithm, and improving the user experience. This mechanism enables the algorithm to process skin image data more accurately, providing users with more personalized, efficient and reliable beauty services.

[0032] For example 2, please refer to Figures 1 to 4 Based on the custom brightness enhancement value TL, after simulating the brightness enhancement of some areas of the face, the average redness of the skin R avg , Average skin glossiness G avg and average skin transparency B avg will change, and then calculate the average redness of the skin after optimization R avg ’ , Average skin glossiness after optimization G avg ’ and average skin transparency after optimization B avg ’ , and the average redness of the skin R avg , Average skin glossiness G avg and average skin transparency B avg The number of noise pixels with large differences from the average pixel in the facial area is calculated as follows: According to HZ R =SQRT(R avg -R avg ’ ) 2 , HZ G =SQRT(G avg -G avg ’ ) 2 and HZ B =SQRT(B avg -B avg ’ ) 2 ) to calculate the redness difference threshold HZ R , gloss difference threshold HZ G and transparency difference threshold HZ B And the skin redness R of the i-th pixel before and after optimization within the total number of pixels N is calculated. i , the skin glossiness G of the i-th pixel i and the skin transparency of the i-th pixel B i , compare one by one, and exceed the redness difference threshold HZ R, gloss difference threshold HZ G and transparency difference threshold HZ B The pixels with large differences are counted as the number of noise pixels with large differences, that is, the number of noise pixels after aggregation XS noise .

[0033] In this embodiment, the average skin redness R in the original image data is avg , Average skin glossiness G avg and average skin transparency B avg It is an important parameter to describe the color characteristics of the facial area of ​​the image. These average values ​​reflect the overall brightness and intensity of the facial skin in different color channels. When the skin simulation module simulates the optimized state of the skin and calculates the difference, it is specifically by comparing the image data before and after optimization. Specifically, the average redness R of the optimized skin in the facial area of ​​the optimized image is calculated. avg ’ , Average skin glossiness after optimization G avg ’ and average skin transparency after optimization B avg ’ , and compared with the original value, the difference can quantitatively describe the impact of the optimization algorithm on the facial skin color characteristics, and this simulation method can also help promote and present the user's beauty methods.

[0034] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A smart beauty method based on facial partitioning, characterized in that: The specific implementation steps are as follows: Step 1: Use the image acquisition module to collect images of the user's facial partitions, transmit them to the image processing module, and store them in the database; Step 2: Using the image processing module, calculate and output the adjusted skin brightness value LF; Step 3: Based on the adjusted skin brightness value LF and the custom brightness boost value TL, the skin simulation module is used to simulate the optimized skin state and summarize the noise pixel number XS. noise ; Step 4: XS based on the number of noise pixels noise and the adjusted skin brightness value LF, and using the image processing module, respectively calculate and output the optimized skin smoothness PF and skin color uniformity value JF; Step 5: Extract the skin color uniformity threshold JF corresponding to the patient's age group from the database HZ , and compare it with the skin color uniformity value JF, and the interactive module outputs the beauty method; The image processing module includes a unit for improving the brightness of the skin in facial zones, a unit for optimizing the smoothness of the skin in facial zones, and a unit for improving the uniformity of the skin in facial zones.

2. The intelligent beauty method based on facial zoning according to claim 1, characterized in that: The image acquisition module includes a high-resolution camera and professional photography equipment; The image processing module includes a computer and image processing software; The skin simulation module includes a skin simulation system; The interaction module includes a display device.

3. The intelligent beauty method based on facial zoning according to claim 2, characterized in that: The calculation formula for improving the skin brightness unit of the facial partition is as follows: <h2 style=";text-align:left;direction:ltr">LF=SQRT(R<h2 style=";text-align:left;direction:ltr"> avg <h2 style=";text-align:left;direction:ltr"> +G<h2 style=";text-align:left;direction:ltr"> avg <h2 style=";text-align:left;direction:ltr"> +B<h2 style=";text-align:left;direction:ltr"> avg <h2 style=";text-align:left;direction:ltr"> )×(1+TL / 100)-Y / 2; Y=YM / FM; in: LF is the adjusted skin brightness value; SQRT is the square root symbol; R avg The average redness of the skin; G avg Average skin glossiness; B avg Average skin transparency; TL is the custom brightness boost value. TL reflects the degree to which the user-defined brightness needs to be boosted. Its value range is {0-100}; Y is the skin shading degree; YM is the shaded area, and FM is the total area of ​​the facial partitions.

4. The intelligent beauty method based on facial zoning according to claim 3, characterized in that: The average redness of the skin R avg The calculation formula is as follows: R avg =(R1+R2+R3+......+R N ) / N; N is the total number of pixels, which reflects the total number of pixels for skin detection in the facial area; R1 is the skin redness of the first pixel, R2 is the skin redness of the second pixel, R3 is the skin redness of the third pixel, R N is the skin redness of the Nth pixel; The average glossiness of the skin G avg The calculation formula is as follows: G avg =(G1+G2+G3+......+G N ) / N; G1 is the skin glossiness of the first pixel, G2 is the skin glossiness of the second pixel, G3 is the skin glossiness of the third pixel, G N is the skin glossiness of the Nth pixel; The average transparency of the skin B avg The calculation formula is as follows: <h2 style=";text-align:left;direction:ltr">B<h2 style=";text-align:left;direction:ltr"> avg <h2 style=";text-align:left;direction:ltr"> =(B1+B2+B3+......+B<h2 style=";text-align:left;direction:ltr"> N <h2 style=";text-align:left;direction:ltr"> ) / N; B1 is the skin transparency of the first pixel, B2 is the skin transparency of the second pixel, B3 is the skin transparency of the third pixel, N The Nth pixel skin transparency.

5. The intelligent beauty method based on facial zoning according to claim 4, characterized in that: Based on the custom brightness enhancement value TL, after simulating brightness enhancement of the face area, the average redness of the skin R avg , Average skin glossiness G avg and skin average transparency B avg will change, and then calculate the average redness of the skin after optimization R avg ’ , Average skin glossiness after optimization G avg ’ and average skin transparency after optimization B avg ’ , and the average redness of the skin R avg , Average skin glossiness G avg and skin average transparency B avg The number of noise pixels with large differences from the average pixel in the facial area is calculated as follows: According to HZ R =SQRT(R avg -R avg ’ ) 2 , HZ G =SQRT(G avg -G avg ’ ) 2 and HZ B =SQRT(B avg -B avg ’ ) 2 ) to calculate the redness difference threshold HZ R , gloss difference threshold HZ G and transparency difference threshold HZ B And the skin redness R of the i-th pixel before and after optimization within the total number of pixels N is calculated. i , the skin glossiness G of the i-th pixel i and the skin transparency of the i-th pixel B i , compare one by one, and exceed the redness difference threshold HZ R , gloss difference threshold HZ G and transparency difference threshold HZ B The pixels with large differences are counted as the number of noise pixels with large differences, that is, the number of noise pixels after aggregation XS noise .

6. The intelligent beauty method based on facial zoning according to claim 5, characterized in that: The calculation formula for optimizing the skin smoothness unit of facial partitions is as follows: PF=LF×(1-XS noise / N)+SQRT(LF×R std ) / 10-XS pore / (LF×2); ； in: PF is the skin smoothness after optimization; XS noise is the number of noise pixels; R std is the standard deviation of skin redness; XS pore is the number of pore pixels.

7. The intelligent beauty method based on facial zoning according to claim 6, characterized in that: The calculation formula for improving the skin uniformity unit of the facial zone is as follows: JF=PF×[1-|R avg -G avg | / (R avg +G avg +1)]+SQRT(LF / (XS pore +1)-PF×(SB / XS pore ); in: JF is the skin color uniformity value; SB is the spot area.

8. The intelligent beauty method based on facial zoning according to claim 7, characterized in that: The adjustment steps of the smart beauty method based on the skin color uniformity value JF are as follows: S1. Determine a custom brightness enhancement value TL according to user needs; S2, calculate the adjusted skin brightness value LF, the optimized skin smoothness PF and the skin color uniformity value JF in sequence; S3, the skin color uniformity value JF and the skin color uniformity threshold JF HZ Make a comparison If the skin color uniformity value JF is higher than the skin color uniformity threshold JF HZ , it reflects that the skin color of the facial area is too uniform and easy to lose the natural feeling. When the pores and spots are reduced and optimized, the spot area SB and the number of pore pixels XS are increased. pore ; If the skin color uniformity value JF is lower than the skin color uniformity threshold JF HZ , it reflects that the texture of the facial partition changes dramatically. When the area of ​​pores and spots is increased and optimized to reduce the spot area SB and the number of pore pixels XS pore ; If the skin color uniformity value JF is equal to the skin color uniformity threshold JF HZ , it reflects that the facial zoning beauty method achieves the effect when the current beauty method is performed.

9. The intelligent beauty method based on facial zoning according to claim 8, characterized in that: The database stores skin color uniformity values ​​JF and skin color uniformity thresholds JF for different age groups. HZ , and regularly adjust the skin color uniformity threshold JF HZ Update, where the skin color uniformity threshold JF HZ Specifically, it is the average value of skin color uniformity value JF of different age groups, and when adjusting and comparing beauty methods, the skin color uniformity threshold JF of the corresponding age group is extracted HZ .

Images