WO2011138649A2

WO2011138649A2 - Method of processing images for application of a colour

Info

Publication number: WO2011138649A2
Application number: PCT/IB2011/000868
Authority: WO
Inventors: Benoît CHAUSSAT; Jean-Marc Robin; Christophe Blanc
Original assignee: Vesalis
Priority date: 2010-05-04
Filing date: 2011-04-18
Publication date: 2011-11-10
Also published as: WO2011138649A3; FR2959853A1; FR2959853B1; EP2593924A2

Abstract

Method of simulating an application of colour to a portion of an image of a face of a user comprising the steps consisting in: - obtaining a photographic image of the user comprising at least one area of application of colour, designed for simulating coloration; - at least for the area of application of colour, extracting with the aid of a module for generating layers, at least one shadow layer and/or one brightness layer suitable for locating and quantifying areas of light and/or of shadow by generating spatial variations of lighting; - at least for the area of application of colour, extracting with the aid of a module for generating layers, at least one colour layer to be applied; - applying, with the aid of a module for applying layers, the layers to the target image.

Description

IMAGE PROCESSING METHOD FOR APPLICATION OF COLOR

TECHNICAL FIELD OF THE INVENTION

The invention relates to an image processing method for simulating a color application on one or more areas of an image of a face while maintaining the realism of the image. The invention furthermore relates to a corresponding method for simulating a color application on a portion of a user's face.

STATE OF THE PRIOR ART

When a grayscale image is colored, the known methods have the effect of adding a color layer to the base image. Typically, a choice is made for a uniform color, which is applied to a portion of the image, without any particular effect. For some applications, this type of neutral coloring is satisfactory. In other cases, the coloring has the effect of reducing or even eliminating the realism of an image. This disadvantage is particularly restrictive in the field of image processing representing human faces, in particular to simulate the application of a colored product or tinted on a face, such as in the field of makeup. For example, to simulate a red application on the lips of an image, a color is added uniformly over the area concerned. The result is a bland mouth, without realism, without relief, etc. In a virtual makeup application, such an image is little or not usable as a model for a user who would like to simulate different color applications, because the result on the image is not comparable with the result of an application of the same color directly on the area concerned (the lips for the previous example).

[0003] Several processes are known today simulating mask productions, for example in the field of make-up. A user provides an image of the face to make up, and gets in return a modified image, on which appears a color mask. This mask is for the user who can use it as a model to obtain makeup. Since it is applied to an image of the user's face and not to an image of a model with different features, the mask gives a realistic effect which is an excellent model for the application of makeup by the user. user or by a makeup artist. It is therefore important that this realistic effect is preserved when color is applied to the image. However, conventionally, known coloring processes have the effect of masking the fine details of a face such as lip textures. SUMMARY OF THE INVENTION

To prevent the image staining steps to be treated, particularly in the field of the simulation of color applications on human face images, give unrealistic results, fades and without relief, the invention provides different technical means.

An object of the invention to provide a method for simulating a color application on one or more areas of an image of a face while maintaining the realism of the image.

Another object is to provide a method for simulating different applications of different colors or shades to make a selection, while providing an optimal resemblance with the object represented by the image, in particular the face.

Finally, another object is to simulate on an image of a face different effects produced by products intended to create variations in brightness or texture or relief. These and other objects are achieved by the method defined in the claims.

In addition, the invention provides an image processing method comprising the steps of: obtaining a target image to be processed; determining a textural layer (or transparency) representative of at least a portion of the textures of this image for an area to be colored; apply to the area to be colored of the target image the textural layer and a coloring layer.

The invention further provides a method of simulating a color application on a portion of an image of a user's face comprising the steps of:

obtaining a photographic image of the user comprising at least one color application zone intended for color simulation;

at least for the color application area, extract using a layer generation module, at least one shadow layer and / or a brightness layer adapted to locate and quantify zones of light and / or shade by generation of spatial lighting variations;

at least for the color application zone, using a layer generation module to obtain at least one color layer to be applied;

-apply, using a layer application module, the layers to the target image.

Such a method is used to generate a layer having the effect of acting on the intensity of the color provided for application, so as to obtain varying intensity levels depending on the position on the application area. In addition, the color to be applied is treated as a layer, and not as a modification or a substitution performed directly on the image considered. It is thus easy to substitute layers, delete or add layers. For example, the color layer to be applied can be replaced by a layer of another color. A layer representing a given texture, independent of the base image, can be added, for example to simulate an effect of sand, moisture, etc. Layer modifications or substitutions can be made for a portion or all of one or more layers.

It is therefore one or more layers of shadow variation and / or brightness variation that can change the lighting level. The generation of spatial variations of lighting (light) makes it possible to reveal the relief of an image. Thus, in an image, to modify a color while maintaining the rendering of the textures, at least one layer of shadow and / or brightness is extracted from it, which will then be applied to the color or colors intended for application. This effectively manipulates color, light and shadows, and therefore textures. It also makes it possible to modify the rendering of a color (brightness, blur, texture) without modifying the fundamental value. We can play on the light and create new effects without distorting colors or textures. This process, which consists of creating layers of "illumination textures", also has the advantage of being able to export the effects created for use on other images.

The method allows to create layers with the ability to locate and quantify areas of light or shade. According to a preferred embodiment, a layer for the light, and a layer for the shadow are produced.

Advantageously, the layers are applied in an order in which there is a lower plane to the upper plane: a target image; the color layer; layers of shadows and / or brightness.

The preferential order of applying the layers allows layers of shadows and / or brightness to act effectively, creating a realistic effect on the resulting image.

[0015] Advantageously, the layers are generated from the histogram of the target image. The inclusion of thresholds, or thresholding, makes it possible to obtain particularly realistic results. According to an advantageous embodiment, the textural layer is a layer of brightness. According to another advantageous embodiment, the textural layer is a layer of shadows. Advantageously, the textural layer comprises a shadow layer and a brightness layer.

Advantageously, the method also provides a step of detecting a plurality of key points of the image to define the area to be colored. In yet another example, an effect layer is also applied, having an effect or texture different from those obtained from the target image. The effect layer is advantageously generated from an image other than the target image.

According to an advantageous embodiment, the layers of a target image having a histogram of N levels, are likely to be obtained by a sum of the partial sums of the pixels of the cards of each of the stages, up to a level. threshold to establish a separation between the levels useful for the determination of a shadow layer, and the levels useful for the determination of a layer of light.

The threshold level of a histogram of a target image is advantageously set so as to separate the pixels into two substantially equal portions. The shadow layer is preferably set from the lowest level to the threshold level. The light layer is preferably set from the highest level to the threshold level.

The invention furthermore provides a device for simulating a color application on a portion of an image of a user's face comprising:

a microprocessor, a working memory and instructions for implementation;

access to image data to be processed;

a layer generation module, adapted for extracting from a target image to be processed, at least one shadow layer and / or a brightness layer adapted to locate and quantify zones of light; and / or shade by generating spatial variations of illumination;

a layer application module, adapted for applying the layers obtained by the layer generation module, to a target image to be processed.

In an advantageous embodiment, the layer generation module is also adapted to obtain, from a color data input, a color layer to be applied.

DESCRIPTION OF THE FIGURES

All the details of embodiment are given in the description which follows, supplemented by Figures 1 to 9, presented solely for purposes of non-limiting examples, in which identical references indicate similar elements, and in which:

FIG 1a schematically shows the operating principle of the method according to the invention;

FIG. 1b schematically shows an example of a device for simulating a color application according to the invention;

FIG. 2 illustrates, from an example, the key steps of the method; FIGS. 3a and 3b show examples of histograms of images or image zones to be treated according to the method of the invention;

FIGS. 4a and 4b show examples of distribution maps of the pixels as a function of the level considered;

FIG. 5 illustrates the principle advantageously used for adding the level maps to obtain the shadow and / or light masks;

FIGS. 6a to 6c show, for purposes of illustration, partial sums of levels, until a shadow layer is obtained;

FIGS. 7a to 7c show, for purposes of illustration, partial sums of levels, until a layer of light is obtained;

FIGS. 8a to 8f show examples of generation of effects of various types;

FIG. 9 shows an example of generation of flake effect.

DETAILED DESCRIPTION OF THE INVENTION

Figure 1b schematically illustrates an exemplary embodiment of a device for simulating a color application on a portion of an image of a user's face. The device comprises a microprocessor, a working memory and implementation instructions, an access to image data to be processed, a layer generation module adapted for extraction from a target image to be processed. , at least one shadow layer and / or a brightness layer adapted to locate and quantify zones of light and / or shadow by generating spatial variations of lighting, a layer application module , adapted for applying the layers obtained by the layer generation module, to a target image to be processed. In this example, the layer generation module is also adapted to obtain, from a color data input, a color layer to be applied. The implementation of the various modules is advantageously carried out by means of implementation instructions, allowing the modules to carry out the operation or operations specifically provided for the module concerned. The instructions may be in the form of one or more software or software modules implemented by one or more microprocessors. The module (s) and / or the software (s) are advantageously provided in a computer program product comprising a recording medium or recording medium that can be used by a computer and comprising a programmed code readable by a computer integrated in said medium or medium, allowing an application software execution on a computer or other device comprising a microprocessor.

Figures 1a and 2 illustrate the operation of the method according to the invention. FIG. 1a is a theoretical schematic representation of the graphic manipulations to be performed to obtain the desired modified image. We observe that a series of layers are superimposed on the to each other in addition to the target image. These layers advantageously include a layer of light, a slap of shadows, a layer of color. The first two types of layers will be described in more detail in the following. The color layer includes the color information provided for application on the target image, and their spatial distribution. The target image is advantageously used in color version (when the original image received is in color) for the application of the layers.

FIG. 2 is a diagrammatic representation of an exemplary implementation of the method according to the invention, from a target image consisting of an image of a mouth on which a user wishes to carry out an application of color optimally, without mitigating the textures of the lips. In step 10, the target image is received and used to generate two layers. In step 20, a shadow layer is produced, and in step 30, a light layer is generated. In step 40, a color layer is obtained. This layer contains the colorimetric data that the user wishes to apply to the target image. The layer 40 can be obtained in various ways, for example by a choice on a color menu, or on images serving as models, or automatically, for example according to a global choice made by the user. than a style or an atmosphere. In step 50, the shadow and light layers are added to each other with the color layer obtained in step 40. The layers 1, 2 and 3 shown in FIG. 2 are preferably asked in order. According to various variants, multiplying factors for one or the other layer can be used, in order to adjust or adapt the influences of the various layers. In step 60, the image processed with the layers is obtained. If the target image to be processed is in color, it is first converted to grayscale. The grayscale version is preferably used to obtain layers. However, as mentioned above, the color version (if it exists) of the target image is preferably used for the layer matching step as shown in Figure 1. In the example illustrated in the figures, 256 levels are considered, in order to obtain a high precision. To minimize the calculations to be made and / or to increase the speed of calculation, for example in a case where limited computing resources are available, the number of levels can be limited, but should be greater than three, and more preferably higher to ten. In uses where precision and realism are decisive factors, the number of levels is advantageously greater than 50 or 100.

Figures 3a and 3b show examples of histograms of target images. These histograms indicate the distribution of the pixels of the image according to their brightness. The abscissa represents the 256 levels. In this figure, the level 0 corresponds to the total darkness (the black) and the level 255 corresponds to the maximum brightness (the white), with a progression of regular preference between the two limits. The ordinate represents the number of pixels assigned to each level. The set of pixels represents the surface of the image. For the processing of a complete image, it is advantageous to use the histogram of the complete image. On the other hand, for the treatment of a portion of an image, such as for example the lips, a histogram corresponding to this portion is advantageously used.

Figures 4a and 4b show examples of pixel maps, for level 30 in Figure 4a and for level 120 in Figure 4b, based on the target image shown in Figure 1. The cards allow to establish a distribution of pixels according to the level.

In the example illustrated in the accompanying figures, the histogram used is that commonly referred to as "RGB", representing all the values of the RGB range, namely red (R), green (V) and blue (B). Using a large number of levels allows for more accurate results, in addition to more pronounced or detailed contrasts. According to an alternative embodiment, the luminosity histogram is used.

The RGB histogram can be represented by the relation: L = (R + V + B) / 3 in which L signifies brightness. The luminosity histogram can be represented by the relation: L = 0.30 R + 0.60 V + 0.10 B.

In Figure 3b, the histogram is split into two sections, separated by a threshold of demarcation. This threshold can be positioned in various places, depending on the cases and characteristics sought. It is advantageously located at the median M of the histogram, so as to distribute substantially 50% of the pixels on each side. On the left side of the demarcation or threshold, levels are taken into account to generate a shadow layer. On the right side of the demarcation, the levels are taken into account to generate a layer of light. Arrow 21 indicates the direction in which levels are taken from level 0 to level 120 in this example. Arrow 31 indicates the direction in which levels are taken from level 255 to level 120 in this example. If a user wants to increase the light level, the threshold can be shifted to the left in this example. If instead a user wants to increase the shadows, the threshold can be shifted to the right. The Median level (M) is preferably not taken into account for the shadow and light layers. This makes it possible to ensure that a level corresponding to a threshold is not influenced by the shadow or the light.

FIG. 5 schematically explains how the maps of different levels are preferentially added together to generate the shadow and / or light layer. At level 1, the level 1 map is used once. At level 2, the level 1 map is added twice, and the level 2 map once. At level 3, the level 1 map is added three times, the level 2 map twice, and the level 3 map once, and so on. The following rule is derived from:

EC = N1 (Map N1) + (N-1) (Map N2) + (N-2) (Map N3) + (N-3) (Map N4) + ... + 1Nseuil (Map Nseuil) where EC is the set of cards cumulated up to level N, N is the number of the current level.

Or, presented differently,

CN1

+ CN1 + CN2

+ CN1 + CN2 + CN3

+ CN1 + CN2 + C N3 + CN4

+ CN1 + CN2 + CN3 + CN4 + .... + CNs

It is the sum of the partial sums evaluated at each level. Stacking all cards on all floors generates the desired layer. To illustrate the process, FIGS. 6a to 6c present, using an image of a face, the intermediate steps before reaching the total sum between the levels 0 and 120. Thus, the result of Partial additions, for example for level 0, the sum for levels 0 to 10, the sum for levels 0 to 20, etc., up to the sum for levels 0 to 120, shown in Figure 6c. There is the gradual appearance of pixels to generate the image to be processed. The shadow layer is then obtained (Figure 6c), with the shaded areas in black and the pale areas in white.

FIGS. 7a to 7c show a similar example, for the addition of the levels 255 to 120. Initially, the level 255 alone, then the sum of the levels 255 to 250, then the sum of the levels 255 to 240, then the sum of the levels 255 to 230, then the sum of the levels 255 to 220, etc., until the sum of the levels 255 to 120. We obtain then the layer of the lights (last image of the figure 7c), with in black, the brightly lit areas, and in white, the dark areas. This mode of presentation of the layer, with the black signifying the presence of information of light, evokes an "inverted" image. In practice, it is not useful to generate all these intermediate steps. Only the final step with the sum of the levels for example 0 to 120 or 255 to 120 is required. This layer of transparency is characterized by a high opacity where the light is maximum and a transparency which increases more the light decreases. Thus, a color applied under this layer has a rendering that takes into account the variations of illuminations. The color is rendered the same where the transparency is strong and undergoes modifications of blur and illumination without being distorted.

In the example corresponding to FIG. 3b, the shadow layer corresponds to the sum of the partial sums (for each stage) between the level 0 and the level 120. The layer of the lights corresponds to the sum of the partial sums (for each floor) between level 255 and level 120. According to one variant, the method is modified by selecting only a fraction of the levels to be stacked instead of stacking the 256 levels. For example, the levels containing only a certain percentage of sufficiently bright pixels can be retained. This makes it possible to create localized effects or to play on the rendering. By decomposing the image and extracting the illumination components, there is thus a way to apply a color in a manner fully adapted to the image. The presented process allows to create a transparent layer that represents the shadows with their different levels of opacity. It is the same for the lights. The shadow layer includes all levels from level 0 to Median level -1, in the case where the threshold corresponds to the median.

To enable the different levels of the intensity of the shadow to be recreated, level 1 is determined as the most shaded level possible, that is to say black (# 000000) and the median level -1 as the less shady. The pixels in the shadow layer are absolute black (# 000000) and have different opacity for each level, represented by an opacity value (Alpha) "Op". To obtain the opacity value corresponding to each level, all the levels represented by the shadows are divided by the maximum value of the opacity, ie 100%.

Op = 100 / Nth

So the opacity of Nmédiane - 1 = Level / 100 and for N1 = Nniveau x (100 / Nniveau) = 100.

To know the opacity of an area of pixels at a given level n, we use the following relation for the layer of shadows:

Op = 1 + (NR - NC) x 100 / Number of Levels

Or :

NR is the level for which you want to know the opacity.

NC is the number of the card for which opacity is desired.

For example, the opacity calculation of the level 2 map 2, out of a total of 120 levels, would give:

Op = 1 + (3 - 2) X 100/120 = 1, 6.

For the layer of lights, the opacity is calculated as follows:

Op = 1 - (NR - NC) x 100 / Number of Levels.

The pixels of the layer of lights are white on a transparent background (Alpha).

According to an advantageous variant of the invention, various types of effects may be applied to the image to be processed. These effects vary the color shades to apply. Moreover, for an application in a virtual make-up simulator, several effects imitating those of certain types of makeup may be provided, in order to widen the possibilities of simulation. Thus, a user can virtually test an almost infinite number of possibilities without having to undergo the constraints of make-up and makeup remover. It can also keep in memory the various simulations performed, for future reference. The effects are presented under different layers representing areas of shadow and light. They are advantageously obtained in a manner similar to the previously described shadow and light layers. In this way, an implementation is obtained which allows a transposition on the target image without loss of realism.

According to an advantageous option, the effect layers can themselves undergo treatments and / or adjustments such as opacity, linear or non-linear blur, coloring, etc. They can also be accumulated, preferably by respecting a certain order.

Below are presented different combinations to obtain certain types of effects such as matte, glossy, creamy, "gloss", etc. These effects are presented as non-limiting examples. A plurality of other effects and / or combinations of effects can be obtained. The matt effect consists of removing from the gloss of the area to be treated, for example the mouth zbiC, as shown in Figure 8a. The goal is to mitigate these brilliances by replacing them with color. As illustrated in FIG. 8a, to obtain a matt effect, an area covering for example 60% of the pixels considered as the gloss set is determined. Depending on the desired effect, a value other than 60% may be used. An es60BC layer represents the addition of the set of levels from 255 to 0 with an identical alpha value for each level. As the demarcation threshold for this example is 60%, the levels taken into account range from about 255 to about 102, so as to obtain about 60% of light. The level addition mode is similar to that previously described for a threshold positioned at the median level (FIG. 2 and FIG. 5).

This layer provides a gray level image zbiCM of the matte mouth. This image serves as a new base for determining a light layer, as previously described. At this stage, the median of the histogram of the zbiCM image is preferably used. To reconstruct the image of the mouth with a matte rendering, then simply superimpose the esMN layers created from the original image zbiC and esMB created from the image zbiCM whose glosses were removed on the zC layer of the desired color. The layers 1, 2 and 3 shown in FIG. 8a are arranged in a manner similar to that presented in step 50 of FIG. 2. The satin effect, shown in FIG. 8b, consists in slightly amplifying the light while at the same time making it soft and diffuse in a defined area. An es60BA layer, obtained in a similar manner to the es60BC layer described for Figure 8a, is obtained from the target image. By adding a Gaussian (or other) type of blur, we obtain the es60BA + EF layer. This layer makes it possible to soften the areas of brightness. It is advantageously applied with an opacity of the order of 50%. The final satin rendering zbiCCS, is obtained by superimposing the esMNA shadow and esMBA light layers created from the original zbiC image whose purpose is to transpose the texture of the lips, with the es60BA + EF layer which allows amplification light. These layers are placed on the layer zC, the layer of the desired color. The layers 1, 2, 3 and 4 shown in FIG. 8b are arranged in a manner similar to that presented in step 50 of FIG. 2.

The brilliant effect, shown in Figure 8c, is to amplify the most intense light already existing in a defined area. From the zone of the target image to be processed zbiC, a threshold covering for example 5% of the pixels is determined. A value other than 5% can be used, depending on the desired effects. The es5BA layer represents the addition of the set of levels from 255 to 0 with no Alpha value (opacity of 100%) until the threshold of 5% is reached. To avoid too much contrast, a slight blur or anti-aliasing can be applied on this layer. The final glossy rendering zbiCCB, is obtained by superimposing the esMNA shadow and esMBA light layers created from the original zbiC image to the es5BA layer that allows the amplification of light, all on the zC layer of the desired color. The layers 1, 2, 3 and 4 shown in FIG. 8c are arranged in a similar manner to that presented in step 50 of FIG. 2. The pearlescent effect, shown in FIG. 8d, consists of amplifying the light on a large area in a delimited area and also to enhance the contrast of the texture. From the zone of the target image to be processed zbiC, a threshold covering for example 70% of the pixels is determined. Another value than 70% can be used, depending on the desired effects. The es70BA is the sum of the set of levels from 255 to 0 with the same Alpha value for each level. The threshold of demarcation being for this example at 70%, the levels taken into account range from 255 to about 76, so as to obtain about 70% of light. The level addition mode is similar to that previously described for a threshold positioned at the median level (FIG. 2 and FIG. 5).

From a second image zbiC + E3, whose contrast has been amplified on strategic areas, another es50BAC layer is created. The final nacreous rendering zbiCCN, is obtained by superimposing the esMNA shadow and esMBA light layers created from the original zbiC image whose purpose is to transpose the original texture, with the layer es70BA whose opacity is low, with the layer es50BAC which allows to accentuate the relief by the light. These layers are superimposed on the zC layer of the desired color. Layers 1, 2, 3, 4 and 5 shown in Fig. 8d are arranged similarly to that shown in step 50 of Fig. 2.

The creamy effect, shown in Figure 8e is to increase the contrast of the texture on strategic areas. zbiC is the area of the image from which we obtain the image zbiC + E3 whose contrast has been amplified on strategic areas. The final zbiCCC creamy rendering is obtained by overlaying the esMNA shadow and esMBA light layers created from the original zbiC image to the es50BA and es50NA layers that enhance the relief. Each group is preferably applied with a certain percentage of opacity so that only the contrast is changed without amplifying the shadows and lights. These layers are superimposed on the zC layer of the desired color. Layers 1, 2, 3, 4 and 5 shown in the figure 8e are arranged similarly to that presented in step 50 of Figure 2. The gloss effect, shown in Figure 8f, is to amplify the light on a defined area giving a gelatin effect on the texture. From the zone to be treated of the target image zbiC, an es15BA module covering approximately 15% of the pixels and another es5BA module covering approximately 5% of the pixels are generated, both being the sum of the set of the included levels of 255 to 0 with an identical Alpha value for each level. The level addition mode is similar to that previously described for a threshold positioned at the median level (FIG. 2 and FIG. 5). On these two modules, a filter is applied, es15BA + E1 and es5BA + E1.

The final gloss rendering zbiCGT, is obtained by superimposing the two layers es15BA + E1 and es5BA + E1 on the original image zbiC for a translucent gloss. This is the variant presented on the right of Figure 8f, with dotted line. For a colored gloss (continuous line on the left of the figure), the same process is used with a coloring in addition to the layer representing the levels going from the median M to 255, producing the layer esMBAC + E1. In both cases the results are directly applied to the zbiC origin area. The layers 1 and 2, or 1, 2 and 3 (as the case may be) shown in FIG. 8f are arranged in a manner similar to that presented in step 50 of FIG. 2. According to another variant of the method according to FIG. invention, the addition of textures or patterns is provided by combining layers of external origin with the layers of the initial image.

For example, it is possible to use layers designed from another image, or any other type of effect layer. This add-on can be added to others or used alone. The zone to be treated may be different from the original one: in this case, it is preferable that the zone on which the levels are taken into account is identical to that of the additional module. For optimal application of this module, the relief of the target zone is transferred to the layer to be added. By way of example, FIG. 9 shows the integration of a layer of flakes. The flakes generally represent points of metallic material that reflect the light. A zpA layer represents them in white (a layer that can be colored in the case of gold or other colored glitter) is created on a transparent background. The flakes will reflect according to the level of light reaching them. The flakes that are exposed to light are therefore considered opaque, and those in the shadow are transparent. The es255NA + esMNA layer of black levels is used to mitigate glitches in shadows. This layer consists of the sum of the levels from 0 to 255 (black threshold) to which the sum of the levels from 0 to the median is added twice, in order to cumulate the effects of thresholds. This layer acts as a mask depending on its opacity. The thinner the opaque threshold, the more the flake has a low opacity. We obtain a layer zpAs which is the additional module to give the effect of flakes. The final rendering zbiCC + zpAs, is obtained by superimposing esMBA, esMNA and zpAs on the zC layer of the desired color. The layers 1, 2 and 3 shown in FIG. 9 are arranged in a manner similar to that presented in step 50 of FIG. 2. The complementary modules can be combined with other complementary modules or effect modules.

The figures and their descriptions made above illustrate the invention rather than limiting it. In particular, the invention and its various variants have just been described in relation to layers applied without correction or adjustment factors. Nevertheless, it is obvious to a person skilled in the art that the invention can be extended to layers applied with various adjustment factors, in order to reduce or increase the influence of certain parameters. For example, most layers can be modulated with opacities ranging from 0 to 100%. The invention is also presented with a particular example in relation to the lips of a face. Nevertheless, it is obvious to a person skilled in the art that the invention can be extended to other areas, particularly human faces or full face. Finally, the invention and its various variants have just been described in relation to layers generated on the basis of histograms comprising 256 levels. Nevertheless, it is obvious to one skilled in the art that the invention can be extended to layers generated on the basis of histograms having a different number of levels. The reference signs in the claims are not limiting in nature. The verbs "understand" and "include" do not exclude the presence of elements other than those listed in the claims. Word

"a" preceding an element does not exclude the presence of a plurality of such elements.

M Median

O Shadow

Light

H Histogram

zC Color area

zbiC area mouth target image

zbiCGT Area Mouth Target Image with Transparent Gloss

zbiCGC area mouth Target image with Gloss Color

zbiCN area mouth image Target Color Nacreous

zbiC area mouth image Target Mat

zbiCCM area mouth image Target Color Matt

zbiCCC area mouth image Target Color Creamy

zbiCCS area mouth image Target Satiny color

esMBA set threshold Median White Alpha

esMNA set threshold Median Black Alpha

es (x) N set threshold (x = percentage of pixels between 0 to 255 of the histogram of the image) Black

es (x) B set threshold (x = percentage of pixels between 255 to 0 in the histogram of the image) White

es (x) BA set threshold (x = percentage of pixels between 255 to 0 of the histogram of the image) White Alpha (white color)

es (x) BAC set threshold (x = percentage of pixels between 255 and 0 of the histogram of the image) Colorized Alpha White (of desired color)

E3 Effect 3 (Nonlinear Filter "Edge Enhancement")

E1 Effect 1 (Nonlinear filter "Median reduced by Alpha")

EF Gaussian Blur Effect

zpA Alpha spangle area

zpAs Alpha flake zone masked by the thresholds

Claims

A method of simulating a color application on a portion of an image of a user's face comprising the steps of:

-apply, using a layer application module, the layers to the target image.

2. A method of simulating a color application according to claim 1, wherein the layers are applied in an order in which is found from the lower plane to the upper plane:

a target image;

-the color layer;

-the layers of shadows and / or brightness.

3. A method of simulating a color application according to one of claims 1 or 2, wherein the layers are generated from the histogram of the target image.

4. A method of simulating a color application according to one of claims 1 to 3, further comprising a step of detecting a plurality of key points of the image to define the area to be colored.

5. A method of simulating a color application according to one of claims 1 to 4, further comprising a step of also applying an effect layer, having an effect or texture different from those obtained from the target image.

A method of simulating a color application according to claim 5, wherein the effect layer is generated from an image other than the target image.

The method of simulating a color application according to one of claims 1 to 6, wherein the layers of a target image having a histogram of N levels are likely to be obtained by a sum of the partial sums of the pixels of the cards of each of the floors up to a threshold level making it possible to establish a separation between the levels useful for the determination of a shadow layer and the levels useful for the determination of a layer of light.

A method of simulating a color application according to claim 7, wherein the threshold level of a histogram of a target image is set to separate the pixels into two substantially equal portions.

The method of simulating a color application according to one of claims 7 or 8, wherein the shadow layer is set from the lowest level towards the threshold level.

A method of simulating a color application according to one of claims 7 to 9, wherein the light layer is set from the highest level towards the threshold level.

Apparatus for simulating a color application on a portion of an image of a user's face comprising:

a microprocessor, a working memory and instructions for implementation;

access to image data to be processed;

12. A device for simulating a color application according to claim 11, wherein the layer generation module is also adapted to obtain, from a color data input, a color layer to a color layer. apply.