FR3118521A1 - Method for determining at least one color compatible with a user's clothing - Google Patents

Abstract

The method for determining at least one color compatible with a user's clothing, comprises: - a step of importing (IMP) at least one photograph image (IM) comprising a representation of at least one part of the wearer's attire; - a segmentation step (SEG) in said at least one image (IM), identifying at least one area of interest (S1, S2, S3) in the clothing; - a step of determining (DET) the color of said at least one area of interest (S1, S2, S3); e t- a step of generating (GEN) said at least one color compatible with the clothing of the user, from the determined color of said zone of interest (S1, S2, S3) as a function of a database of compatible colors (PAL1-PAL4) or a calculation rule for obtaining compatible colors (RGL, 11-71). Figure for abstract: Fig 2

Description

Method for determining at least one color compatible with a user's clothing

Modes of implementation and embodiments relate to a determination of one or more color(s) compatible with a user's clothing, for example in the context of assistance or recommendation in a choice of a makeup cosmetic product.

The term “compatible” color with the clothing is understood to mean a color which is in an aesthetic agreement, which presents a harmony, with the colors of the clothing.

There are services, exercised by people qualified in color compatibility ("make-up artists" or "make-up artists", usually called "make-up artists"), to recommend compatible color combinations, i.e. that is to say aesthetic colors between them according to criteria accepted by consensus, subjective criteria or criteria relating to fashion phenomena.

Because of the subjective aspect of this type of recommendation, it is difficult to offer automated techniques for determining colors compatible with an outfit, without the outfit being known in advance.

However, there is a need to provide compatible color recommendations to make-up consumers who do not have make-up artist qualifications, as a function of a given outfit varying with each use, and immediately (that is to say say without consulting a person qualified in the matter by appointment).

The recommendations of compatible colors can thus serve as an aid to make-up consumers to make an aesthetic choice, or even serve to obtain a product having this color compatible with the clothing.

Indeed, new “connected” methods of consuming cosmetic products are emerging, and allow for example to obtain from home a cosmetic product whose recipe is unique and manufactured by a device controlled from a multifunction mobile phone. This is particularly the case of the cosmetic product dispenser described in the US patent application filed on December 31, 2020 under number 17/139,338, and in the US patent application filed on December 31, 2020 under number 17/139,340.

Thus, a determination of a color compatible with a given outfit each day can be used to generate the cosmetic product having precisely this compatible color, instantly and daily.

In this respect, according to one aspect, a method is proposed for the automatic determination, within a computer system, of at least one color of a cosmetic product, for example a cosmetic make-up product, compatible with an outfit of a user, comprising:
- a step of importing at least one photograph image comprising a representation of at least part of the user's clothing;
- a segmentation step in said at least one image, identifying at least one area of interest in the clothing;
- a step of determining the color of said at least one area of interest; and
- a step of generating said at least one color compatible with the clothing of the user, from the determined color of said zone of interest according to a database of compatible colors or a calculation rule obtaining compatible colors.

The computer system can for example be a single user computer device, such as a multifunction telephone, a touch pad or a computer, or else a user computer device and an external computer server capable of communicating and cooperating together.

The method according to this aspect can thus be implemented in an automated manner, for example by means of a simple user computer device such as a multifunction telephone, to determine colors compatible with an outfit, without the outfit being known in advance.

The process can thus provide the user with one or more colors for their make-up depending on their clothes, and possibly their preferences.

In addition, the method can make it possible to recommend purchases of complete make-up products thanks to the association of colors, or else to order an "ideal" recipe for each outfit, manufactured by a personal device of the type of the cosmetic product distributor connected mentioned above.

According to one mode of implementation, the generation step made according to a database of compatible colors, comprises:
- a step of identifying, among color banks containing a finite number of colors chosen beforehand to be mutually compatible, the color bank containing the closest color to the color of said determined zone of interest; and
- A step of selecting said at least one color compatible with the clothing from among the colors of the identified bank.

According to one mode of implementation, the generation step made according to a calculation rule for obtaining compatible colors, comprises:
- a calculation step comprising a conversion of the color of said determined zone of interest into a first point of a suitable colorimetric space, and an application of at least one mathematical transformation of this point, said at least one mathematical transformation being previously chosen to result in at least a second point of said space whose color is compatible with the color of the first point; and
- a step of selecting said at least one color compatible with the clothing from among the color(s) of said at least one second point.

According to one mode of implementation, the selection step comprises a selection of said at least one compatible color from among the colors obtained, belonging to a family of colors selected by the user.

The color family may correspond to a user preference, or else to a family of possible recipes according to cartridges loaded in a personal device of the type of the connected cosmetic product dispenser mentioned above.

According to one mode of implementation, the generation step is implemented with a machine learning model.

Advantageously, the machine learning model is configured to adapt the selection step according to previous choices of the user on compatible colors determined by previous implementations of the method.

The machine learning model can for example be of the convolutional neural network type, and makes it possible to benefit from extremely fine precision in the generation of the color compatible with the clothing, which is particularly advantageous in this context of harmony of colors and aesthetics.

According to one mode of implementation, the segmentation step is implemented automatically with a machine learning model.

The machine learning model can here again be of the convolutional neural network type, and on the one hand makes it possible to avoid segmentation errors of a non-useful part (for example a part of the body or of the background) , and in addition to detect advantageous details of the clothing outfit that are too delicate to obtain by techniques without machine learning, such as colors of accessories and jewels or small patterns on the outfit.

According to one mode of implementation, the segmentation step comprises a request addressed to the user to identify the position of said at least one area of interest on the image.

According to one mode of implementation, the step of determining comprises a colorimetric processing of said at least one image with a machine learning model adapted to correct a deformation of the color of the representation of said at least part of the outfit user's clothing due to the lighting conditions of the photograph and/or the device that captured the photograph.

For example, the machine learning model adapted to correct color distortion may be of the type described in the French patent application filed on February 3, 2021 under number FR2101039.

According to another aspect, a computer system is also proposed, for example being a single user computer device or else a user computer device and an external computer server capable of communicating and cooperating together, intended to determine at least one color of a cosmetic product compatible with a user's clothing, comprising:
- communication means configured to import at least one photograph image comprising a representation of at least part of the user's clothing;
- processing means configured for:
- perform a segmentation in said at least one image so as to identify at least one area of interest in the clothing;
- determining the color of said at least one area of interest; and
-- generating said at least one color compatible with the clothing of the user, from the determined color of said zone of interest according to a database of compatible colors or a calculation rule for obtaining compatible colors.

According to one embodiment, the processing means are further configured to implement the segmentation steps, the determination step and the generation step of the method as defined above.

According to one embodiment, the system includes a user computer device configured to capture or store said at least one photograph image in memory and to communicate said at least one photograph image to the processing means during said import.

According to another aspect, a computer program product is also proposed comprising instructions which, when the program is executed by a computer, lead the latter to implement the method as defined above.

According to another aspect, a computer-readable medium is also proposed comprising instructions which, when executed by a computer, lead the latter to implement the method as defined above.

Other advantages and characteristics of the invention will appear on examination of the detailed description of modes of implementation and embodiments, in no way limiting, and of the appended drawings, in which:

;

illustrate embodiments and embodiments of the invention.

The illustrates an example of implementation of a method 100 for determining at least one color CC compatible with a user's clothing.

The method 100 is for example intended to be implemented by a computer device, such as a multifunction telephone (usually “smartphone”), a touch pad, or a computer. Optionally, the method 100 can be implemented in collaboration with an external server in a configuration where the server receives input data (IM image), executes at least some of the operations of the steps of the method 100 (among SEG, DET, GEN, SEL), and provides output data (CC).

Thus, the method 100 can be materialized, in practice, by a computer program product comprising instructions which, when the program is executed by a computer, lead the latter to implement the method 100.

The method 100 can also be materialized, in practice, in the form of a computer-readable medium comprising instructions which, when executed by a computer, lead it to implement the method 100.

The method 100 for determining at least one color CC compatible with an outfit of a user IM, can for example be exploited as a service offered to the user to provide him with recommendations of colors, for example of make-up, compatible from an aesthetic point of view according to a given outfit varying with each use, and immediately (that is to say without consulting a qualified make-up artist by appointment).

The recommendations of compatible colors can thus be used to assist make-up consumers who do not have the qualification of make-up artists, to make an aesthetic choice, or possibly serve to obtain one or more commercial product(s) having the color or colors (s) compatible with the clothing.

Indeed, the method 100 can also be used to recommend purchases of complete make-up products thanks to the association of colors, or else to order an “ideal” recipe for each outfit, manufactured by a personal device of the type of connected cosmetic product distributor.

In this regard, the method 100 comprises on the one hand a step of importing IMP of at least one photography image IM comprising a representation of at least part of the clothing of the user.

For example, the IMP import of the photography image can result from a capture made by the user, in self-portrait mode or possibly by photographing his reflection in a mirror.

Then, the method 100 includes a segmentation step SEG in said at least one IM image, so as to identify at least one area of interest S1, S2, S3 in the clothing.

The SEG segmentation step can be implemented automatically by an image segmentation algorithm, for example adapted to perform object recognitions, contour definitions, pattern recognitions, background suppression, or other suitable and known image processing.

In particular, the SEG segmentation step is suitable for identifying elements, called “areas of interest”, S1, S2, S3 of the clothing present in the IM image.

The segmentation step SEG is for example capable of detecting an element S1 of the upper part of the clothing, close to the face, such as a blouse, a t-shirt or a bodysuit.

The segmentation step SEG is for example capable of detecting an element S2 of the middle part of the clothing, at the level of the pelvis, such as pants, a skirt or a dress.

The segmentation step SEG is for example capable of detecting an element S3 of the lower part of the clothing, close to the feet, such as shoes, socks or the bottom of trousers, a skirt or a dress.

Of course, the elements S1, S2, S3 identified in the image depend on the shooting of the photo, and for example shoes may not appear in the IM image.

In the example illustrated by the , the middle zone of interest S2 and the lower zone of interest S3 both correspond to part of a long skirt. In such a situation, the segmentation SEG is for example configured to identify and isolate a pattern and remove the background color from the skirt in the area of interest S2, and to identify and isolate the background color from the skirt and remove the pattern in the area of interest S3.

Furthermore, the SEG segmentation step can be implemented automatically with a conventional machine learning model, for example of the convolutional neural network type. This makes it possible on the one hand to avoid segmentation errors of a non-useful part (for example a part of the body or of the background), and on the other hand to detect advantageous details of the clothing which are too delicate to obtain by techniques without machine learning, such as accessory and jewelry colors or small patterns on the outfit.

Alternatively, as will be described below in connection with the , the segmentation step SEG can be performed “manually” by the user, for example by means of a request addressed to the user to identify the position of said at least one zone of interest on the image IM.

From the areas of interest S1, S2, S3 identified during the segmentation step SEG, the method 100 includes a step DET for determining the color of the areas of interest S1, S2, S3.

The determination of the color DET can for example extract the dominant color from each zone of interest S1, S2, S3, or possibly make an average of the colors present in each zone of interest S1, S2, S3.

Advantageously, the DET determination step includes colorimetric processing of the IM image, suitable for correcting color distortion in the image caused by the lighting conditions of the photograph and/or by the device having captured the photograph.

Indeed, the color may be distorted depending on the nature of the device that captured the photograph. There are methods to compensate for the deformation relative to a given material.

On the other hand, lighting conditions can also alter the perception of a color, typically the color will have a colder color temperature if the lighting conditions, i.e. the light sources in the scene photographed, are cold; the color will have a warmer color temperature if the lighting conditions are warm.

A machine learning model can advantageously be provided to correct color distortion, for example by taking into account the geographical position of the shot (indoor or outdoor), the timestamp of the shot (day or night ), the weather at the time of the shot (sun, clouds, rain, snow). The parameters used to estimate the correction to be made can be more complex than the basic elements presented above.

The machine learning model can for example be of the convolutional neural network type. The use of a machine learning model makes it possible to benefit from an extremely fine precision in the DET determination of the color of the areas of interest S1, S2, S3 of the user's clothing, which is particularly advantageous in this context of color harmony and aesthetics.

After the determination DET of the “real” colors of the areas of interest S1, S2, S3, the method 100 comprises a generation step GEN of at least one color CC compatible with the clothing of the user.

The generation GEN is made from the “real” color(s) determined in the determination step DET.

In this regard, two approaches are possible ID-SEL, RGL-SEL.

The first RGL-SEL approach can use "the seven rules of the science of color and harmony" to target the formation of a certain type of relationship between the colors of clothing and compatible colors.

In this first RGL-SEL approach, the GEN generation step is made according to an RGL calculation rule for obtaining compatible colors, and includes an RGL calculation step followed by an SEL selection step.

The RGL calculation step comprises a conversion of the color of the area of interest determined into a first point "dep" of a suitable colorimetric space, for example the "HSV" space, and an application of at least one mathematical transformation 11, 21, 31, 41, 51, 61, 71 of this “starting” point dep.

The mathematical transformation is previously chosen from among the seven rules of science and harmony, to result in at least a second point of said space whose color is compatible with the color of the first point dep.

We define the seven rules of science and harmony, in the “HSV” or “Hue-Saturation-Value” colorimetric space well suited to the characterization of colors by humans, as follows:

The first rule 11 is a monochromatic transformation, that is, a transformation of different saturations and values of the same hue as the starting color dep.

The second rule 21 is a complementary transformation to the starting color dep, i.e. a projection towards the hue diametrically opposite to the hue of the starting color in a representation of the hues in a disk of revolution.

The third rule 31 is an adjacent complementary transformation, i.e. a projection towards the neighboring hues of the hue complementary to the starting color dep.

The fourth rule 41 is a transformation into adjacent colors, that is to say a projection of hues close to the hue of the starting color dep.

The fifth rule 51 is a quadratic transformation, that is to say a projection of three hues forming, with the starting hue dep, a square centered on the disk of revolution of the hues.

The sixth rule 61 is a triadic transformation, that is to say a projection of two hues forming, with the starting hue dep, an isosceles triangle centered on the disk of revolution of the hues.

The seventh rule 71 is a rectangular transformation, that is to say a projection of three hues forming, with the starting hue dep, a rectangle centered on the disk of revolution of the hues.

Advantageously, in all the rules 21-71 except the first monochromatic rule 11, the saturation and the value of the starting color dep will be kept in the colors obtained by the various transformations 21-71.

From the colors obtained by the various transformations 11-71, the generation GEN of the CC-compatible color comprises a selection SEL of at least one CC-compatible color (j.i-k) from among the resulting color(s) .

For example, the resulting colors can be selected according to their belonging to a color family, chosen by the user. For example, in the context of make-up of the foundation or lipstick type, the color families can be “Les Rouges”, “Les Oranges”, “Les Fuchsias” or “Les Nus”.

The color family can correspond to a user preference PREF ( ), or else to a family of possible recipes according to cartridges loaded in a personal device of the connected cosmetic product dispenser type.

Furthermore, the GEN generation step can be implemented with an adapted and trained machine learning model to select the colors resulting from the seven rules of color science and harmony best suited to a type of makeup. given.

Further, the machine learning model is advantageously configured to adapt the SEL selection step based on prior user choices on compatible colors determined by prior implementations of method 100.

Here again, the machine learning model can for example be of the convolutional neural network type. Again, the use of a machine learning model allows extremely fine precision in generating the DC-compatible color with the attire, which is particularly advantageous in this context of color harmony and of aesthetics.

The second ID-SEL approach can use a predetermined palette based on a makeup artist's recommendation for a seasonal style in combination with the color of the attire.

We refer in this respect to the .

The illustrates the method 100 within the framework of the second RGL-SEL approach of the generation step GEN of the compatible colors CC.

In this second RGL-SEL approach, the GEN generation step is made according to a PALi compatible color database, 1≤i≤4, includes an ID identification step followed by an SEL selection step.

Banks, or “palettes”, of colors PAL1, PAL2, PAL3, PAL4, each containing a finite number of colors are defined beforehand by a make-up artist so as to contain colors that are aesthetically compatible with one another.

For example, each PALi palette, 1≤i≤4, can correspond to a seasonal style, "winter", "spring", "summer", "autumn".

The identification step ID ( ) comprises an identification, among the color banks PAL1, PAL2, PAL3, PAL4, of the color bank PALi, 1≤i≤4, containing the closest color to the color of said determined area of interest S1, S2, S3.

From the colors obtained by the various transformations 11-71, the generation GEN of the CC-compatible color comprises a selection SEL of at least one CC-compatible color (j.i-k, for example j=3, i=2, k= 1 or 2 or 3) among the colors of the bank(s) identified PALi, 1≤i≤4.

For example, the resulting colors can be selected according to their belonging to a family of colors FAMj, 1≤j≤4, chosen by the user PREF. For example, in the context of make-up of the foundation or lipstick type, the color families can be “Les Rouges”, “Les Oranges”, “Les Fuchsias” or “Les Nus”.

The color family FAMj, 1≤j≤4, can correspond to a preference of the user PREF, or else to a family of possible recipes according to cartridges loaded in a personal device of the connected cosmetic product dispenser type.

Thus, a CC-compatible color can be identified by a couple j.i obtained in a table with a line "i" corresponding to the palette identified PALi, 1≤i≤4, and a column "j" corresponding to the chosen family FAMj, 1≤d≤4. Several CC-compatible colors can be selected for the same pair j.i, each being then identified by an additional index “k”. For example the three compatible colors 3.2-1, 3.2-2, 3.2-3, i.e. j=3, i=2, k=1 or 2 or 3, are obtained from the FAM3 family "Les Fuchsias" in the PAL2 "Spring" palette.

When several palettes PALi, 1≤i≤4, are identified for respectively several areas of interest S1, S2, S3, then priority is advantageously given to the area of interest S1 closest to the face of the user on the IM picture.

Consequently, for a chosen family FAMj, j=J, if a single palette PALi, i=I, is identified, then the selection step SEL provides K compatible colors J.I-k, 1≤k≤K, for example K =3.

For a chosen family FAMj, j=J, if two palettes PALi, i=I1 or I2, are identified, then the selection step SEL provides K1 compatible colors J1.I1-k, 1≤k≤K1; and K2 compatible colors J1.I2-k, 1≤k≤K2, with K1+K2=K and K2≤K1, for example K1=2 and K2=1.

For a chosen family FAMj, j=J, if three palettes PALi, i=I1 or I2 or I3, are identified, then the selection step SEL provides K1 compatible colors J1.I1-k, 1≤k≤K1; K2 compatible colors J1.I2-k, 1≤k≤K2; and K3 compatible colors J1.I3-k, 1≤k≤K3, with K1+K2+K3=K and K3≤K2≤K1, for example K1=1, K2=1 and K3=1.

In addition, the generation step can be implemented with a machine learning model adapted and trained to generate the colors belonging to the same PALi palettes, and the best suited to a given type of makeup.

Further, the machine learning model is advantageously configured to adapt the SEL selection step based on prior user choices on compatible colors determined by prior implementations of method 100.

Here again, the machine learning model can for example be of the convolutional neural network type. Again, the use of a machine learning model allows extremely fine precision in generating the DC-compatible color with the attire, which is particularly advantageous in this context of color harmony and of aesthetics.

Furthermore, as an alternative to the automated SEG segmentation described in relation to the , the illustrates an example of SEG segmentation done “manually” by the user.

Indeed, the user can position probes S1, S2, S3 on the areas of interest of his choice in the clothing, for example thanks to a touch screen displaying the IM image. From the point of view of the method 100, for example executed by a multifunction telephone SMPH, the segmentation step SEG comprises a request addressed to the user to identify the position of said at least one area of interest S1, S2, S3 on the IM picture.

Thus, the illustrates a computer system SYS intended to determine at least one color compatible with a user's clothing, adapted to implement the method 100 described above.

The system SYS includes a user computer device, for example the multifunction telephone SMPH, and possibly an external computer server.

The user computer device SMPH includes communication means configured to perform the import step IMP of the method 100, for example via the internet network and with the communication means of the computer server.

For example, the user computing device SMPH typically comprises a photographic sensor CAM configured to capture the IM photography image, and/or typically a memory suitable for storing said at least one IM photography image.

The external computer server, or alternatively the user computer device SMPH, comprises processing means configured to perform the segmentation step SEG of the method 100, the determination step DET of the method 100, and the generation step GEN CC compatible colors.

When the processing means are all incorporated in the user computing device SMPH, the import step corresponds to an internal transfer of the data of the image IM, from the photographic sensor CAM or from the memory, to the processing means .

In other words, there has been described a method 100 and a corresponding system which, using the CAM camera of an SMPH telephone, extracts the color of certain places ("areas of interest") from clothing, such as a blouse, pants or shoes.

With this set of colors, Process 100 then classifies each of them into a specific "universe" or "season" (Summer, Winter, Spring, Fall) by finding the closest color match in the color banks. pre-existing PAL1-PAL4 and their corresponding color worlds.

In addition, a "sub-selection" is made in a family of colors FAMj desired PREF by the user to additionally specify the hues of the colors.

Make-up products can then be recommended in the same PALi universe and possibly the same FAMj family of colors, giving priority to the colors closest to the face, in order to optimize the coordination of the user's make-up with his outfit.

The choice of the FAMj family can correspond to the colors that can be produced by a connected make-up dispenser, based on the set of cartridges installed in the dispenser, also on the basis of the number and type of cartridges of the dispensing device , and based on the number and priority of probes the user has decided to use. The user can also choose another FAMj family to browse the options that would be available with other cartridge sets. This may prompt the user to buy a new set of cartridges.

In addition, the use of artificial intelligence makes it possible to offer the user a new type of color recommendation service. By taking a more holistic approach by including other elements of their daily attire, such as their clothes, shoes, and accessories, recommendations can be adapted frequently based on user preferences.

Claims (14)

Method for automatically determining, within a computer system, at least one color of a cosmetic product compatible with a user's clothing, comprising:
- a step of importing (IMP) at least one photograph image (IM) comprising a representation of at least part of the user's clothing;
- a segmentation step (SEG) in said at least one image (IM), identifying at least one area of interest (S1, S2, S3) in the clothing;
- a step (DET) for determining the color of said at least one area of interest (S1, S2, S3); and
- a generation step (GEN) of said at least one color compatible with the clothing of the user, from the determined color of said zone of interest (S1, S2, S3) according to a base of compatible color data (PAL1-PAL4) or a calculation rule for obtaining compatible colors (RGL, 11-71). Method according to claim 1, in which the generation step (GEN) made according to a database of compatible colors (PAL1-PAL4), comprises:
- an identification step (ID), among color banks (PAL1-PAL4) containing a finite number of colors chosen beforehand to be mutually compatible, of the color bank (PAL1) containing the closest color of the color of said determined area of interest (S1, S2, S3); and
- a selection step (SEL) of said at least one color (c3.1, c3.2, c3.3) compatible with the clothing from among the colors of the identified bank (PAL1). Method according to claim 1, in which the generation step (GEN) made according to a calculation rule for obtaining compatible colors (RGL, 11-71), comprises:
- a calculation step (RGL) comprising a conversion of the color of said determined zone of interest (S1, S2, S3) into a first point of a suitable colorimetric space, and an application of at least one mathematical transformation ( 11-71) of this point, said at least one mathematical transformation (11-71) being chosen beforehand to result in at least one second point of said space whose color is compatible with the color of the first point; and
- a selection step (SEL) of said at least one color (c3.1, c3.2, c3.3) compatible with the clothing from among the color(s) of said at least one second point. Method according to one of Claims 2 or 3, in which the selection step (SEL) comprises a selection of the said at least one compatible color from among the colors obtained, belonging to a family of colors (FAMa-FAMd) selected by the user. Method according to one of the preceding claims, in which the generation step (GEN) is implemented with a machine learning model (AI_GEN). Method according to claim 5 taken in combination with one of claims 2 to 4, in which the machine learning model (AI_GEN) is configured to adapt the selection step (SEL) according to previous choices of the user on compatible colors determined by prior implementations of the method. Method according to one of Claims 1 to 6, in which the segmentation step (SEG) is implemented automatically with a machine learning model (AI_TCOL). Method according to one of Claims 1 to 6, in which the segmentation step (SEG) comprises a request addressed to the user to identify the position of the said at least one zone of interest on the image. Method according to one of the preceding claims, in which the determination step (DET) comprises a colorimetric processing (TCOL) of the said at least one image (IMk) with a machine learning model (AI_TCOL) suitable for correcting a distortion the color of the representation of said at least part of the user's clothing due to the lighting conditions of the photograph and/or to the device (CAM) having captured the photograph. Computer system intended to determine at least one color of a cosmetic product compatible with a user's clothing, comprising:
- communication means (COM) configured to import at least one photograph image (IM) comprising a representation of at least part of the user's clothing;
- processing means (PU) configured for:
-- performing a segmentation (SEG) in said at least one image (IM) so as to identify at least one zone of interest (S1, S2, S3) in the clothing;
-- determining (DET) the color of said at least one area of interest (S1, S2, S3); and
-- generating (GEN) said at least one color compatible with the clothing of the user, from the determined color of said zone of interest (S1, S2, S3) according to a color database compatible colors (PAL1-PAL4) or a calculation rule for obtaining compatible colors (RGL, 11-71). System according to claim 10, in which the processing means (PU) are further configured to implement the segmentation steps (SEG), the determination step (DET) and the generation step (GEN) according to one of claims 2 to 9. System according to one of Claims 10 or 11, comprising a user computing device (APP) configured to capture or store in memory said at least one photograph image (IM) and to communicate to the processing means (PU) said at least one photograph image during said import (IMP). Computer program product comprising instructions which, when the program is executed by a computer, lead the latter to implement the method according to one of Claims 1 to 9. Computer-readable medium comprising instructions which, when executed by a computer, lead the latter to implement the method according to one of Claims 1 to 9.