FR2783949A1 - Remote digital processing of an image, in particular an image of a human face - Google Patents

Abstract

The image processing system extracts characteristic features of a human face (for example hair, lips etc.) and by detection, analysis and recognition of detail characteristics and of texture. A number of tools are provided for manual selection and adjustment of the image, and for applying small transformations under pre-set rules to selected regions.

Description

The present invention relates to a computer and digital tele system.

  intelligent to transform virtually, automatically, almost

  instantly and very precisely the initial image (s) of a real target, for example a human face and certain of its characteristics (hair, lips etc.), by detection, analysis and recognition of its fine characteristics and its respective textures, supplemented by a help and correction system, drawing tools, retouching of photographic images, coloring incorporating color effects of automated textures, supplemented by a virtual selection tool that will allow to apply a meticulous transformation, according to established rules, in the games

  previously identified and cut out.

  We know that there exist in computer graphics techniques, and in particular in the art of makeup, a certain number of manual or automated drawing and coloring equipments. This equipment makes it possible to modify the initial image of the face of an individual and at the same time, and to deduce therefrom manually or automatically a randomly aesthetic image, even imprecise for the application of makeup drawings on the previously identified surfaces and clipped. Also, existing processes only color by restoring colors laid flat on the image, without seeking to particularize them by their original texture aspect when this is the case. These systems can neither restore colors in their original texture aspect, nor apply them in the context of an image (light, shadow, etc.). Virtually transformed images are bland, with no relief effect. This handicap considerably harms the sales or commercial effect of a product, for example in the art of make-up, lipsticks which use several textures, such as gloss, mat, etc., or certain hair coloring products with shiny texture or with multiple reflections. These individual processes are more or less rapid and require the use of heavy and specialized fixed equipment which is in the form of a cabin or a console equipped accordingly, and which can be used particularly, or even exclusively by necessity on a specific location (store). The size, the difficulties of maintenance, of distribution (languages, places, accessibility, supports) and the cost price of such equipment do not allow a widely popularized use, even domestic of the device in the whole world immediately. If the detection, analysis or recognition fails, there is no intelligent system to help and correct these operations. There is also no automatic generation of a personalized color / product palette in harmony with a subject's face. Still in this process, the existing devices do not allow a subject, when he obtains the virtual transformation of his face, for example for a makeup or a hair coloring, to choose, to

2 2783949

  almost instantly select and modify the aesthetic characteristics of its new virtual image using a virtual selection tool integrating a palette of harmony and custom. Finally, the restitution of colors in their real aspects and textures (shine, matte, powdery, reflections etc.) is an operation which requires the use of many computer graphics and algorithmic tools which performed manually, is practically impossible in a very short time. short, and involves several delicate and tedious operations, or even very random precision depending on

the skill of the skilled person.

  Finally, access to data is limited by the local storage capacity of a cabin

or the computer system.

  The present invention therefore relates to a method and a device enabling several thousand, even several million users, anywhere in the world, at the same time and in a few seconds, to translate virtually, by microcomputer, tele computer means. and digital models, the representation and visualization of their new images that is as precise and realistic as possible aesthetically in the expression they wish to give themselves, from an initial target,

a face for example.

  The invention can be used from home or at any other suitable place in the world. It does not require the technical complexity of a cabin and access to data

is no longer limited.

  To help and correct the detection, analysis and recognition of a face, by

example.

  Automatically restore colors to their original texture aspects

  for spectacular and truly realistic virtual effects.

  To offer personalized palettes in harmony with the characteristics of his face. To choose, select and modify almost instantly the aesthetic characteristics of your new virtual image using a virtual selection tool integrating

  a palette of harmony and personalized.

  The device used is, in its first destination, in the form

  a digital multimedia and TV terminal or console (dressing table, for example)

  communicating. In order to miniaturize the device as much as possible, all the databases as well as the archiving system can be stored on a remote Internet server which is accessed via the Internet. It essentially comprises an Intel 486 DX2 or Pentium or equivalent microprocessor, of 60Mhz power and higher, coupled with a 32MB RAM memory and a 4MB video memory, with or without:

3 2783949

  a floppy drive / recorder, or a hard disk or a Digital Virtual Disk (DVD) or any other means of digital storage, for example an electronic hard disk memory (hardship disk), of at least 100 megabytes of capacity, at least one CCD digital video sensor of the type used in video communication coupled to its video interface or a color scanner, a reflecting or semi-reflecting mirror with 50% refraction power or not, a color touch screen or not of at least 35 cm diagonal screen size, can be plasma technology or active matrix liquid crystal with its specific video / graphics interface allowing a minimum VGA display resolution of 640 x 480 dots and 256 colors, even 1280 x 1024 dots and a depth of 24-bit colors (16.7 million colors) or more, a keyboard and its mouse, or a remote control (infrared or radio frequency), or a voice recognition system (control by voice), a modem with a speed of at least, 6Kbits PSTN or digital ISDN type, a color printer type liquid or solid inkjet, or thermal sublimation, or laser or any other means of editing colors, a language called Transmission Control Protocol and Internet Protocol (TCP / IP), union of processes forming a basis for data transfers over the Internet, navigation software such as Netscape, Internet explorer etc., a set of capture tools digital images, an automatic recognition engine configured to be interfaced and interactively interacted with the class 0+ or 1 expert system and its database (s), and an editing module that will generate the results obtained, in the form of '' a color plate of the initial real image and the new virtual image and certain comments relating to artistic or technical criteria in the

  area in which this image could have been created.

  This tele-communicating multimedia terminal is connected in a transparent manner to one or more Internet servers which are computers on the Internet or World Wide Web network which is a device, on the scale of the Internet network allowing, to generate a fabric of very publications. variety with an increasing number of multimedia and intelligent functions. These servers have the resources of the TCP / IP language, the

  description used for the presentation of information (HTML) from the World Wide Web,

  the HTTP (Hyper Text Transfer Protocol) protocol for data transmission, from one or more class 1 expert systems equipped with one or more databases, and driven by an inference engine, i.e. - say a powerful deductive system independent of the database, and coloring and drawing tools for the transformation. This expert system is interactively interfaced with the recognition tool, that is to say that it is capable of processing information in its own right.

4 2783949

  provenance and to send him back the orders and the tools he deducted for the

transformation of the image.

  Conversely, this device implemented as described above, can be modulated according to the case and the circumstances. Thus, it can take two other different usage configurations, without hardware modification, and distinct:

  In its second destination, it takes the form of a very simplified terminal,

  that is to say that he does not have any software tool apart from a system allowing him to access and browse the Internet, he is permanently or on demand connected via the Internet to one or more Internet servers which will allow him to use a set of digital image capture tools, an automatic recognition engine configured to be interfaced and interactively interacted with the class 1 expert system and its database, and a results editing module that will generate the results obtained, in the form of a color plate of the initial real image and the new virtual image and certain comments relating to the artistic or technical criteria of the

  or areas where this image may have been created.

  In its third destination, it takes the form of an autonomous intelligent terminal, that is to say that it has all the integrated software tools, that is to say that it has in memory a system allowing it to '' access and browse the Internet, a set of digital image capture tools, a detection, recognition and automatic analysis process interfaced interactively with its class 0+ or 1 expert system, and its database, and an editing module that will generate the results obtained, in the form of a color plate of the initial real image and the new virtual image and certain comments relating to artistic criteria or

  techniques of the field or fields in which this image could have been created.

  In its first destination, the present invention therefore relates to a process which can be distributed and accessible everywhere in the world, and by several thousands or even several millions of users at the same time, from a space-saving tele-computer and multimedia process and at very low cost. cost, making it possible to automatically transform, by an automated and rapid process of piloting drawing tools, photographic retouching and coloring, the initial real image of a target, for example a face, into a new image, but virtual, generated through a choice of possible combinations deduced from the expression in which one wishes to be virtualized, characterized in that the initial and real image, captured at least by a digital color CCD video sensor of the type used in video communication coupled with its video interface or a color scanner, is analyzed and recognized in its context, for example as u n human face, segmented, then in that we determine the whole

2783949

  corresponding fine characteristics of this image by recognizing shapes, zones, colors, dimensions, textures and contours, and finally that some of these characteristics are modified, according to the resources available through Internet servers distant, using decision criteria from one or more expert systems, provided with one or more knowledge bases enriched with multiple combinations and harmonies corresponding to an expression in one or more fields which will then allow the tools to coloring, photographic retouching and drawings to automatically and digitally modify the points of the initial image, in the zones and contours previously detected and adequate, corresponding to the transformation of the initial and real image into a new virtual image in an expression given. To allow if it is the case, to automatically restore the colors in their aspects of original textures in a precise context, for spectacular and truly realistic virtual effects. To offer personalized palettes of colors in harmony with the characteristics of a subject's face. To leave the possibility to a user to modify, almost instantly the image of this newly simulated face, using a selection tool which offers him, among a personalized range in perfect harmony with this face, the choice to modify and '' try the colors plus their textures, and the products of one or more manufacturers at the same time and in different activities, for example make-up and

hairdressing.

  The console is connected to the server resources using a modem with a speed of at least 55.6Kbits PSTN or digital ISDN type or another telecommunication method (radio frequency etc.), d '' an Internet address which geoinformatically locates the specialized server site for this or these operation (s) in the WEB, a language called Transmission Control Protocol and Internet Protocol (TCP / IP), union of processes forming a basis for data transfers on the Internet , a software of

  navigation such as Netscape, Internet explorer etc ...

  As soon as the terminal is powered up, it becomes active, ie it is ready to transmit its requests and receive resources from one or more identified servers or that the user indicates to it. In a hypermedia environment, the tools will allow the capture of the color image which will be obtained in a few seconds at

  minimum means of a digital color CCD video sensor of the type used in video

  communication coupled to its digital color video interface coded at least 16 bits or from the color scanner, in good conditions for shooting and lighting

appropriate or not.

6 2783949

  In its second destination, the terminal connects to one or more dedicated Internet servers on which the tools and data are available. This connection is made by means of a modem with a speed of at least 55.6Kbits PSTN or digital ISDN type or any other means of telecommunication (radiofrequency etc.), from an Internet address which locates geoinformatically the server site suitable for this operation in the WEB, a language called Transmission Control Protocol and Intemrnet Protocol (TCP / IP), union of processes forming a basis for data transfers on the Internet, navigation software such as Netscape, Internet explorer etc. . In its third destination, the terminal works locally without a permanent Internet connection. The terminal has locally all the shooting, detection, recognition, analysis, help and correction tools, a system

  class 0+ or 1 expert with its database and a results editing module.

  The terminal connects to one or more Internet servers to carry out

  updating or cleaning or enriching its local database.

  It is from an image, previously recognized, a face for example, sized and segmented by specific zones, that the method and the device according to the invention will allow in its context to transform it into the desired expression. The process essentially comprises four phases whatever the destination of the terminal: An acquisition phase: recognition, analysis, detection of areas, shapes, textures and contours of a color photographic image are carried out. digitized using deductive techniques,

algorithmic and deductive.

  - A transformation phase: we extract the data from the

  acquisition phase which are transmitted to the class () + or 1 expert system as appropriate.

  The data is then processed by the class O + or 1 expert system according to the decision criteria established in a rule base and the information contained in the database or knowledge, which will lead to the transformation of the characteristics of the 'initial image. Certain colors depending on the context will be analyzed and recaractified according to a particular texture (shiny, matte, powdered etc.). The results from processing by the class 1 expert system will then be retrieved and processed using the drawing, segmentation and recognition tools.

shapes.

  - A restitution phase or virtual "flash" proposal: the initial image will be transformed on the terminal screen or the console mirror, which will lead to

7 2783949

  get the new virtual image in the expression we wanted. This image is

displayed in color.

  - A phase of games and final choices: the process displays on the screen a virtual selection tool composed of a personalized palette which is essentially characterized by a range in perfect harmony with one's face for example (colors, products, objects, accessories). Thus the user will be able to modify the simulated virtual image, by

  testing the harmonic and personalized choices offered to him on the screen.

  This operation makes it possible to modify the zones of the image automatically in a few seconds, and will lead the user, either to accept the image resulting from the virtual "flash" proposal, or the new one which he will have reconstructed using virtual tools available. The final image and its technical characteristics are then edited on

  A4 paper for example, locally or on another site and on all available supports.

  During the first phase, the user will photograph or digitally record the color image of his face which will be displayed in a framing test pattern, provided for this purpose on the screen, via the shooting tool. The image will be automatically flipped to be right, i.e. the left side of the face image is on the left, and the right side of this same image is on the right. This operation which lasts only a few seconds will lead, in a second step to

  automatically activate image processing tools.

  During the second phase, the image processing tools will coordinate several successive operations: First, it detects that the initial image is present and

  correctly framed in the test pattern.

  Secondly, it detects and recognizes areas and

contours.

  When the tool has determined the top and bottom of the face, as well as the sides, it can, thanks to precise dimensional criteria, determine the gender, or even the type at which

  belongs the face (woman, man, adult, child.).

  To this end, the histograms of the 256 (16.7 million) colors are considered according to the display and resolution power of the video card, coded on 16 or 24 bits, fitted to the terminal, corresponding to each elementary segment of the image. Following two privileged directions (vertical and horizontal for example), it determines the points where the histograms undergo a discontinuity and the points where the histograms pass through minima or maxima so as to optimize the method, the derivatives of the variance will be determined , the mean and the coefficient of variation of the color levels of each image line. We thus arrive at precisely determining the

8 2783949

  zones, it detects and recognizes zones and contours. The color image of the face thus detected will be segmented horizontally into six random zones. So the tool will detect the top and bottom of the face, as well as the left side and

the right side.

  From these elements, the tool will delimit (segment) the image at least in six horizontal zones starting from the top and to the bottom of the face, and in two vertical zones: For the horizontal zones Zone 1: the hair (upper face) Zone 2: forehead Zone 3: eyes Zone 4: lower eyes to tip of nose which implicitly includes cheeks and cheekbones Zone 5: tip of nose to upper lip Zone 6: lower lip up to the tip of the chin (lower face) For vertical zones Zone 1: left part of the face Zone 2: right part of the face Then, this operation will allow him to position the different elements of the face by compared to others, to determine the shape and color of the characteristic facial features (nose, eyebrows, eyes, eye color, lips etc.) and

  generate "geometric primitives".

  Objects can thus be easily isolated from the background. The regions of interest of the image are delimited. The tool can then extract the characteristic features and automatically cut and sharpen them, then extract the colors of each object by

  example the color of the eyes, that of the skin.

  During this operation, the tool can call on automatic contour following machines or, to optimize its layout and its detection, virtual "templates" models. These virtual "templates" models will in this case be the "master standards" for the recognition and extraction of the contours of the object sought. These virtual "templates" models have statistical information on the shape and size of the object in question. That is to say that we have in memory, statistical and forecast coordinates of the drawing of the reference contour of each element or object targeted, for example a lip

9 2783949

  superior. The statistical data of the outline drawing that he will have extracted, must then be correlated with that of the drawing of the reference shape, of the model.

corresponding virtual "template".

  The observation of a face leads to theoretical reasoning, that all the elementary forms which characterize it are curved. When the shape or the context of the zone are delicate, in particular when the outline drawing happens to be outside the statistical limits of the outline drawing of a virtual "template" model, for example when it is about bad elements specified in a face (lips of a color very close to that of the skin for example), the outline drawing is obtained, by calculating and mathematically applying a parabola on the element beforehand and approximately recognized in one or more segments from which we will statistically deviate the curve, point by point, according to the average and the coefficient of variation of the color levels of each line of this zone, until each of the points draws with precision and finesse the contour. The newly obtained outline drawing is compared again to the virtual "template" model of the object or

  the corresponding element until perfect correlation.

  In all cases, the operation which consists in segmenting or extracting and finely trimming the shapes of the targeted objects contained in the image, generates the drawing of the vertical and horizontal lines, for the segmentation, in the form of a green net and also for drawing the shapes of objects, the outline drawing appears as a net

  green. They are displayed over the initial image displayed on the terminal screen.

  The interest of this operation allows the user to follow and participate fully in this

  phase, or even interrupt it if it fails.

  In the event of an exception, both detection and recognition may have failed, we use

  then the help and correction tool.

  We use the help and correction tool for detection, analysis and pattern recognition for certain cases, in particular to increase or reduce the surface area of the face to be analyzed, or in case of detection failed. The user will be able to operate and correct manually. The subject will then specify the location of the face in the image (detection of the face image) with the recognition tool, and force him to inspect and re-analyze the zones or targets in this face image.

until the success of the operation.

  For this operation, the subject uses a manual tool to control the facial framing lines. The vertical and horizontal lines will be moved in a precise direction, to be positioned until framing edge to edge, and correctly the face, that is to say the top of the face (top of the hair), bottom of the

1 (2783949

  face (tip of the chin), left and right sides of the face. Line shifting operations are carried out by means of a computer peripheral such as the tactile parts of the screen or the cursor keys on the keyboard or the ball of the mouse or the remote control or the voice. Then, it can relaunch the detection, recognition and extraction of the characteristics of the image of this face. In case of additional difficulty during detection and recognition, for example if the hair merges with the complexion of the face (very blonde type), we can operate according to two treatments: During the first treatment, the horizontal and / or vertical lines will be positioned on the face image. By this operation, we will reduce the surface area of the face to be analyzed. In this case, the user must position the horizontal line from the top to the extreme limit of the hair and the skin of the forehead, and no longer at the top of the hair. The subject can then launch the detection, recognition and extraction of

  characteristics of the image of this face.

  During the second treatment, at the end of the above operation, the subject will correct the area to be analyzed by repositioning the top line this time at the crown of the hair (origin), which will implicitly determine an area between the hair and the extreme limit of the forehead skin (former positioning). Thus the hair will be detected, recognized and analyzed. We can complete and reinforce this intervention by

  also using the two vertical lines.

  It is perfectly conceivable that in the application of the process, the tool described below

  above, can very well be used from the start of the process when shooting (digital photography in the target sight). Thus, instead of launching the process fully automatic, the subject can launch it in a semi-automatic mode. This mode can be selected according to several options: menu, function keys, voice command, etc. It fully uses the functions of the help tool. This mode can be used for example in case of doubt about the visual aspect of the image, the physiognomy of the face, the

  background etc., or in case of voluntary limitation of the area to be analyzed.

  Thirdly, the recognition and analysis of skin textures

  follow pattern recognition.

  We evaluate the degree of entropy of the skin texture which takes into account the importance of the organization or disorder of this tissue as well as the existence

  or the absence of spatially repeating patterns.

  The presence of skin dislocations corresponding to

  scars, traces of rosacea, which may appear on the image.

  It also reveals the presence of artificial colors corresponding to a make-up or a coloring of the hair which, in this case, will be corrected by an operation of elimination of these colors in order to obtain the image of a virtually cleansed face or d '' hair in its natural color while retaining texture effects (shine, etc.). The tool also analyzes the dominant colors which appear in the different regions of the face (skin color, eye color, hair color, eyebrow color etc.), these criteria thus extracted will then make it possible to determine with accuracy the right colors, in a range by

  example, and in perfect harmony with this face.

  This phase also includes a determination of the structural effects during which we generally assess the degree of homeomorphism of the tissues, the degree of image symmetry, and more particularly the degree of orthogonal symmetry.

  when the object examined is a face (chirality).

  In all these treatments, the image quality can be improved by eliminating or attenuating noise, and by analyzing and correcting light, contrast, gamma, and focus. This quality can also be improved by the use of special effects applied to the image (artistic blurring, modification of the background, integration of the image into another environment etc.) These transformations can be contextual, i.e. -to say that the parameters of the operators are likely to vary according to an environment or a position with respect to the frame of the image or with respect to characteristic shapes previously detected.

  The tool improves image quality by automatically applying filters.

  Filters should be classified into two categories:

  - The integration operator filters likely to attenuate the noises.

  - Differentiation operator filters which, on the contrary, can enhance contrast, modify light, adjust gamma or focus, etc. Depending on the case, these filters are applied locally or globally on the image and they must possibly take into account the local context of the image, taking into account

its statistical properties.

  These filters are intended for operations to be carried out in at least 256 colors.

  In all cases, the application of these filters requires management of the intermediate "caches". During the transformation phase, the image recognition tools communicate all the data they have extracted, for analysis and processing

12 2783949

  by the expert system of class () + or 1. The expert system will then determine the

  corrections and modifications to be made to the image in the theme chosen by the user.

  In the Internet environment, the class 0+ or I expert system may reinforce part of its processing towards another class () + or 1 expert system located on another Internet server site and managing additional data. The results are then

  automatically transmitted for processing by the image recognition tool.

  During the transformation phase of the original image, the process provides the

  improvements or effects that have been previously desired by the user.

  So for example, if it is a make-up operation, the traces of a previous make-up are eliminated, the eyes, the cheeks, the lips, the skin of the face

  will be redesigned based on the new colors and designs.

  If they are small wrinkles, they will be greatly reduced. etc. Overall, that is, depending on the image detection, recognition and analysis tool supplemented by the decisions of the expert system (s), and its database (s), fine facial characteristics are automatically modified and redefined automatically according to harmonic and personalized choices. During this phase, all of the software tools automatically and according to the rules of the art of the make-up profession, provide aesthetic improvements in

different regions of the face.

  In the case where the data necessary for expert system inferences may be incomplete or uncertain, the use of fuzzy logic makes it possible to weight the premises of the rules and the rules themselves so as to be able to infer even in the presence of incomplete data or imprecise. Cosmetic choices relating to make-up can thus be made even if we do not have precise

all the necessary data.

  In the decision phase, the following operations are also deducted. E Colors supplemented with their original textures, for example for make-up are analyzed, calculated and defined in their particular context (lipsticks or "Gloss" or powders etc.). The tools will then apply according to the texture of the area (lip, cheek, hair etc.) the color corresponding to the makeup, but also and transparently the effect of the cosmetic product, that is to say that the 'we are going to reproduce its real appearance, for example its shine, its powdery or shiny glittery appearance, its matt appearance etc. This operation takes into account the context of the image in each of its zones (lighting, brightness, shadows, reflections, etc.), which will allow using algolithmic tools to calculate the texture, to define it in its real appearance, such

13 2783949

  that it would be reproduced in reality. By this process, the quality and the realistic properties of the image are very significantly improved. Thus, during the restitution, one avoids putting the makeup colors in the form of a flat, which harms the conformity and the realistic, artistic and aesthetic realization of the makeup. By this operation, the colors and the finely reproduced make-up drawings fully respect the objectives of the art of make-up, bringing light, relief and realism to the new

virtual face image.

  The colors in harmony with the characteristics of the face are taken into account.

  The expert system determines a color palette among 256 (or more, up to 16.7 million) colors available, correlated with those of a range available in its available database (s), based on data from the recognition tool. Thus, during the restitution phase, the tool will be able to make a coloring proposal in harmony with the face for example, but also offer a selection of colors, from a range, in perfect harmony with this face. In the same way, the expert system will determine the seasonal trend that characterizes the subject. The trends can be of four types: summer, spring, winter, or fall, and each of its trends corresponds to appropriate colors, which

  allows the system to determine more precise harmonies.

  The expert system prepares the virtual selection tool which integrates a personalized palette of choice from colors, products, accessories, objects, etc. Thus, during the restitution phase, the user can automatically and freely modify the new virtual image which has been proposed to him by the tool. In the art of make-up, this personalized palette can take the form of three selection lines, for example: Lips: 6 colors in brown tones with different textures (matte, shiny, etc.) plus a choice of use of a pencil for the outline of the lips The eyes: 6 colors in blue and purple tones with duets (2 colors at the same time degraded on the eyelid) The cheeks: 6 blushes in chestnuts with different textures and aspects (transparent, dark, light, powdery, shiny, glittery, reflections, etc.) The expert system will also prepare the drawing tools that will be used to apply certain colors to the face. For example in the art of makeup, the eyes can be painted in the form of cat eyes, that is to say that the colors placed on the outside of each eye must stretch upwards as a wing ending in a point. The lips can also be painted but also redrawn using the

color and pencil.

14 2783949

  These drawings are specific and they are defined beforehand with "a person skilled in the art". The expert system has in the database the appropriate drawings and forms which are characterized as being "virtual templates" previously defined and which will be recalculated and readjusted to the areas of the face to which they operate, according to the information of the tool. recognition, the context of the image, and the effect they should suggest. In the invention, for more understanding, we call this

  automatic drawing and coloring tools.

  The restitution phase first results in an aesthetic proposal characterized by the transformation of the initial image into its new virtual image on which appear in perfect harmony the desired aesthetic modifications and the expression that one has desired (for example sophisticated make-up, complemented with a coiled hairstyle and a hair color). This results in a display in minimum 256 colors or 16.7 million or more, depending on the power of the video / graphic interface and the quality of resolution and the size of the color screen. The makeup products as well as their corresponding colors can also be viewed, or even supplemented by the virtual selection tool which integrates a personalized palette of choice from colors, products, accessories, objects etc., with which the user will be able to play in view of modifying almost instantaneously certain zones (lips, eyes, cheekbones, whole face, hair, ears for earrings etc.) in the representation of his new

  virtual image until his final decision.

  The final result can then be edited locally, on an inkjet, laser or thermal sublimation color printer, in A4 format or any other technically available format. The content of its information formulates a beauty prescription, including the initial image and the transformed image, advice, facial characteristics (shape, eye color, hair, etc.), product photos, color palette. personal harmonies, clothing color advice in relation to our colors, etc. The results can be similarly edited on the high definition printers of the Internet server which will take care of sending it by post to the client. These same results can be translated, in the same way, on or in different media, preprinted or not (CV, virtual postcard, multimedia clip, video, calendar, banner, small poster, photo album, etc.) available through the applications of the server. They can be stored in the terminal's magnetic or electronic memory or on the Internet server for later use. The

  new image with or without information can be sent by the E- function

2783949

  mail and using the "Attach file" command to one or more correspondents

  with an email address.

  Still in the application of the process to the human face, the expert system is able to simulate, on request, styles of make-up adapted to the circumstances and his personality. We can thus make a make-up best suited to a party, an evening, for the work, express an image in relation to your community or your own criteria

creativity.

  The expert system is able to simulate on demand and check the effects

of a skin care program.

  The implementation method in the present invention will be better understood by

  referring to the description of a simplified example concerning a woman's face described in

  reference to the accompanying drawings, in which: - Figure 1 represents the target sight intended for the digital shooting of the

face.

  - Figure 2 shows the image of a woman's face to be modified, taken from the shot in its framing pattern, in 256 colors or more Figure 3 represents the woman's face in Figure 2 recognized by the face recognition automaton and the extracted data transmitted to the expert system, and shows in green lines, the areas of the image subject to recognition and contour plots, to the location of the color of the iris of the eye (+), when locating the wings of the nose (it), - Figure 4 represents the face modified by the process of the present invention including the colors, the shapes, the designs, the textures of the products, the coloring of the hair, the hairstyle and accessorization, - Figure 5 shows the face modified by the process with its virtual selection tool, and - Figure 6 shows an example of the transformation of the image of a woman's face. Figure 1 is the sight in which the subject will correctly place his face, in view of the digital photographic shooting by the tool. The subject will position his face so that the vertical line identified (A) must be placed

  long and in the middle of the bridge of the nose, and the eyes should be respectively on the one hand

  above the horizontal line marked (B) and on the other hand each eye must be located

  positioned in each of the frames marked (AB) and (AB ').

16 2783949

  FIG. 2 is the image of a woman to be modified obtained in the form of a digital shot using a color video CCD sensor in the target sight in 256 colors

  or more which will allow computer processing of this image.

  FIG. 3 represents the computer image of this face in color, formed of elementary segments, delimiting six horizontal zones and two vertical zones, each characterized by a histogram specifying for this elementary segment the level of each of these colors. Having isolated the forms from the backgrounds, we were able to trace

  automatically the outline of the targeted objects.

  The analysis and recognition of the shapes of this face have been made possible: We have chosen two preferred directions such as the lines (V) and (H) in Figure 3. Following these preferred directions, we determine the points such that (V1; Hi) o the histograms undergo a discontinuity corresponding to the limit of the dark hair of the face studied. We will also reveal all the points such as (V1; H2) or (V1

  ; H5) o the histograms pass through minima.

  By moving the lines V and H along the image of the face, we define lines

  continuous representing the shape of the different elements of the face.

  We are then able to recognize and analyze the skin textures by examining, for example, the histograms of the points noted according to the

  directions such as H2, H3 ..... H7.

  Having determined the shapes and nature of the skin textures in this image, we are able to move on to the phase of extracting the contours and analyzing the colors. In this case, the tool analyzes the color of the eyes, eyelashes, eyebrows, eyelids,

  of the facial skin, then the coloring of the lips.

  Having defined the elements to be modified (objects: eyelids, eyes, lips, cheekbones

  etc.), we extract the contours of these elements.

  Having made automatic photographic corrections to improve the quality of the image: elimination of noise (parasitic points), brightness, color balance, contrast, focus, gamma etc. We therefore memorize all the points of each of these zones or targets to modify their color and redraw their shape to make up the makeup in the illusion

that we wanted.

  In the event of recognition failure due, for example, to poor shooting, etc. With the help of the help and correction tool for detection, analysis and recognition, the user will ability to operate and correct manually. The subject then goes

17 2783949

  specify the location of the face in the image with the recognition tool, and force it to re-inspect the areas or targets in this image of the face until it

  recognize and determine them correctly.

  This intervention and the resulting coTrrection will operate as follows: By displaying a frame in the image with four lines marked (V), (H), (V2) and (H8). These lines are movable, i.e. the subject can move them freely in

precise direction.

  By using and dragging the lines (V), (H), (V2) and (H8), the subject frames the image of the face. The lines are then adjusted edge to edge with the parts of the face: upper face (H), lower face (H8), left side of the face (V), right side of the face (V2). For certain exceptional features, the lines (H) and (H8) can be repositioned, for example if the hair is significantly confused with the color of the skin (a very blond subject), in this case the line (H ) up to the limit of the hair and the rest of the face (limit determined by the line (Hi) during recognition). In this case, the hair can not be recognized during the first recognition, to do this we must restart a second recognition by dragging the line (H), this time at the top of the face (top of the hair), the line ( H1) having been determined during the first recognition ((H) was equal to (Hi)). The recognition tool can then determine the shape and texture of the hair. We can complete and optimize the operation by also using lines (V) and (V2) by

  readjusting manually, at the limit of the hair relative to the bottom.

  FIG. 4 corresponds to the restitution phase during which the fine characteristic features of the initial image have been transformed to give a completely new and aesthetic image, following deductions from the expert system and the

color texture.

  Thus we note that the image represents in figure 2, a woman with the erased aspect,

  hair gathered back, without parting, the face is neither made up nor accessorized.

  In this restitution phase (Figure 4), a certain number of modifications made it possible to modify this face very significantly: The image of the woman's face is turned upside down. The background of the image has been slightly corrected to blue. The image was changed by makeup, hairstyle and color, and accessories. The dermatological defect (red acne pimple) has been reduced. Depending on the hair color (dark brown lightened towards the tips), the shape and color of the eyes and eyebrows, the eye area has been redesigned. Thus, the eyelids were colored, with a slightly red / pink and ivory duo, from the inside

  towards their outer ends.

  The drawing of the coloring of the eyes ends slightly in a point, outwards, according to a slight oblique pulling gently upwards, with the obvious aim of making the look more feline. The tool used a dual make-up, to improve the contrast with the color of the woman's blue eyes, slightly red / pink (on the bottom of the eyelid) and slightly ivory (on the top of the eyelid). The eyes are underlined with a light pencil stroke

  intense black, on the upper part, to emphasize the look with the rest of the face.

  Depending on the shape of the face (round in this case) the tool chose to put a blush in porcelain tones, very back and on each side of the face,

  choosing to stretch it from top to bottom, in order to ovalize the face (effect of refining the features).

  The face was completely covered with a very light porcelain foundation.

  The lips have been redesigned with a discreet pale pink lipstick including a medium shine effect (color texture). it is noted that the restitution of the textures of the makeup products makes it possible to obtain more demonstrative effects. Thus the tool allows to better highlight certain areas such as the lips which have a beautiful

curve effect.

  The hairstyle was modified by stretching and smoothing the hair. The hair was colored using intense natural red tones. This color was applied transparently

  respecting and thus including, the texture of the original hair, in particular the reflections.

  For the accessories, the tool chose artificial plants which finish

boldly adorn the face.

  FIG. 5 shows the virtual selection tool, representing, for makeup for example, four zones or targets identified by the design of the eye, lips, cheeks / skin, as well as corresponding palettes for the colors and their textures (shiny, matte, etc.), extracted from a makeup product range for example, and in perfect harmony with this face. This operation will allow the subject to select and modify by the expert system the zones or targets of this image, almost instantly by means of a computer peripheral such as the keyboard, or the tactile zones of the screen, or the mouse, or the remote control, or voice recognition (voice). To do this, it will indicate to the tool by means of one of the computer peripherals, the zone or target, the lips for example and select a neutral or textured color in the corresponding palette. Once this operation is complete, the recognition tool modifies the zones or targets of the face image. The selection can be made zone or target by zone or target

19 2783949

  one at a time, or two at a time, or three at a time, and so on until you determine

yourself your new image.

  The virtual selection tool, if programmed for this purpose, can offer, in addition to the colors of makeup or hair coloring, the colors of clothing that will harmonize perfectly with the whole. It is this automatically generated image that reveals her new face to the client. This is what figure 6 represents: the image on the left represents the front face

  transformation, and the image on the right after transformation.

  All of their information can formulate a beauty prescription, or a proposal including the initial image and the transformed image, advice, facial characteristics (shape, eye color, hair, etc.), product photos. , the personal color palette of our harmonies, a clothing color advice in relation to our colors etc. The results can be edited in the same way on the high-definition remote printers of the Internet server which will be responsible for sending them to the postal address of his client. These same results can be translated, in the same way, on or in different pre-printed media or not (CV, virtual postcard, multimedia clip, video, calendar, banner, small poster, photo album, etc.) available through the server applications. They can be stored in the terminal's magnetic or electronic memory or on the Internet server for later use. The new image, whether or not completed with information, can be sent by the E-mail function and using the "Attach file" command to one or

  several correspondents with an e-mail address.

  This process and this device can obviously be applied to the simulation of a virtual image from a color photograph previously scanned using a high definition color scanner (1200 X 600 ppi for example). This operation will make it possible to process this image by computer and to make, when it is a photographic portrait of a subject, certain corrections or improvements. Thus, we can proceed to the aesthetic improvement of this face using hair techniques and make-up, blurring or rectification of certain anatomical defects, change of skin color (tan, bleaching etc.), in in this case, the database corresponding to the aesthetic simulation of the face is enriched by a collection of aesthetic rules concerning makeup, hairstyling, hair techniques, skin texture and / or plastic surgery allowing simultaneous treatment the face and the capillary part of a subject to edit or reproduce very quickly a new more rewarding image, see ideal thanks in part to the automation and to the precision of the process, but also thanks to the perfect restitution of colors and their textures (shiny, glittery

2783949

  shiny, matte, powdery, reflection etc.) in a range of makeup or hair products. The visual effect is all the more realistic and artistic as when setting the colors we take into account both the texture of the target area (lip for example), the light of the image, and the texture of the color itself. -Even which will be placed transparently on its target. It is easy to see that this system can receive many applications in

  supplementing the expert system with commercial or scientific data.

  As a sales aid and animation tool in the aesthetic field The device according to the invention makes it possible to help make a better choice in the products of a range of a manufacturer (hair or cosmetic) and to provide truly personalized aesthetic advice, very quickly depending on the personality of one or more brands. So if you are in a domestic environment, store, or even the Internet, a brand may very well offer virtual makeup or hairstyling services using the tool to its customers. We can thus order a personalized make-up and / or a hairstyle. The device then simulates the modifications to be made, as well as the future appearance of the face, and the products can be ordered in

online or on site.

  This image will allow to realize the model that we wished for upon receipt of its products. The invention applied in the Intemrnet field is all the more interesting as it offers virtual tools for make-up, hairstyling, etc., in an immaterial manner, but even more precise than the person skilled in the art in reality, and responds to a request well

  real and natural in this area by users.

* The expert system database is enriched with rules concerning make-up techniques for those skilled in the art and make-up products or

  aesthetic and artistic techniques of a manufacturer.

  In plastic surgery We can thus visualize, from the existing anatomy of the face to be treated, the modifications to be made such as a facelift, a correction of the nose, a maxillary transformation, a correction of the hare's beak, the removal of scars , hair implant etc. The software then simulates the changes to be made, as well as the future appearance of the face. The expert system database is enriched with rules concerning anatomy

  and the corresponding surgical technique.

21 2783949

  The invention in this field makes it possible to obtain a rapid and precise diagnosis of

  operations to be carried out, even to reserve on the Internet, his clinic and his surgeon.

  After healing, this last image must be compared with the result obtained, in order to see

  if there is still reason to make some additional corrections.

  In dermatology The use of an expert system for processing the cutaneous image in color allows a new development by quantifying and qualifying the condition of the skin thanks to

  texture and color distribution analyzers.

  In particular, the image acquisition and processing system associated with an expert system makes it possible to carry out a diagnostic aid model and to suggest the nature of the treatment to be carried out (ointment, active care cream, etc.). , the device makes it easier for the dermatologist to appreciate the evolution of the condition and to easily observe the color changes of

  skin tissue (eczema, shingles, psoriasis, dermatitis etc.).

  The database of the expert system is enriched with information concerning the changes in color of the skin tissue or skin spots corresponding to a set of conditions, accompanied by examples of treatments, making it possible to suggest

  the nature of the treatment to be carried out.

  Hairdressing and hair technique The device according to the invention makes it possible to help make the choice of a better aesthetic by facilitating the modeling of hairstyle and / or of different hair colors. Since this application can be implemented directly by the method, the application and the capillary techniques can be directly included in the

facial treatment.

  The database corresponding to the aesthetic simulation of the face is completed by a set of aesthetic rules concerning the hairstyle and the color of the hair thus making it possible to simultaneously treat the face and the capillary part of a subject. Jewelery and ornamental accessories for the face The device according to the invention helps to make the choice of one or more accessories or jewelry by facilitating the search according to the manufacturer (s), tastes and price , then to try them virtually, in all hygiene, even without pain if it is by

example of piercing accessories.

22 2783949

  This application can be implemented directly by the process, the application, the products and the accessorization techniques can be directly

  included in image processing.

  The database corresponding to the aesthetic simulation of the face is completed by a set of aesthetic rules concerning accessories for facial ornaments and jewelry, thus making it possible to simultaneously treat the face and the part

ornament of a subject.

  In eyewear The device according to the invention makes it possible to help the eyewear optician to help make the choice of a better aesthetic by facilitating the superposition of different spectacle frames on the appropriate area of the face. Also to modify the color of his eyes, by modifying the color of the iris of the eye by imitating the available colors of the lenses.

  color change contact.

  Since this application can be implemented directly by the method, application to the technique of eyewear or optical instruments can be directly included in the treatment of the face. The changes in color of the iris can react on the deductions of the expert system in makeup for example, or even in the ornament of the face for the choice of colors. Same thing, the deduction of

  some frames can be made with a new suitable hairstyle.

  The database corresponding to the aesthetic simulation of the face is supplemented by a collection of aesthetic rules concerning spectacle frames, optical instruments (coloring lenses, lenses as an advertising medium, etc.) allowing thus to simultaneously treat the face and superimpose

  different glasses frames adequately or different colored lenses.

  In physical or virtual access control for example on a network (for example

Internet) or a computer site.

  The system makes it possible to recognize the face of a known person without fail, which can make it possible to trigger an opening system, for example an airlock to access the vault of a bank, or to open his rights users on a network or

computer site.

  The database is enriched with information corresponding to all the characteristic features (shape, dimensions, colors, etc.) of one or more known faces, making it possible to quickly and surely recognize the face of a so-called person.

  authorized and grant him his rights.

In identity photography

23 2783949

  The system according to the invention allows an identity photograph to improve

  the aesthetics and the image of a subject.

  The database corresponding to the aesthetic simulation of the face is completed by a collection of aesthetic rules concerning make-up, hairstyling, hair techniques, eyewear, facial ornamentation, and the texture of the skin, thus making it possible to treat simultaneously. the face, the hair part of a subject for

  edit photographs that are rewarding for the client.

  In photographic processing The system according to the invention makes it possible, from a color photograph, on all supports (Argentine paper, slide, special paper for digital printing etc.) and in all formats (identity, portrait etc.), after having been "scanned" by a camera

  appropriate scanning, to be retouched and improved aesthetically.

  This process applied to mail order sales is a good way to drive sales of a brand, especially in cosmetics. Customers provide their photographs on an appropriate medium provided by the service provider. The brand restores them some time after a new image of them aesthetically in line

  with their own sales targets and their own.

  Then if the client authorizes it, his image can be archived to be used for

  personalized recovery campaigns either by post or on the Internet.

  The database corresponding to the aesthetic simulation of the face is completed by a collection of aesthetic rules concerning make-up, hairstyling, hair techniques, eyewear, facial ornamentation, and the texture of the skin, thus making it possible to treat simultaneously. the face, the hair part of a subject for

  edit photographs that are rewarding for the client.

  In morphopsvchology and modeling of individual profile The device according to the invention makes it possible to observe and "read" with meticulousness a face, wrinkles and fine lines (their situation, their directions and their depth), the balance between the upper part and below, the size and shape of the nose, the ear lobe and the corners of the lips, the implantation of the hair, the size of the forehead relative to the rest of the face etc., thus it is possible to detect and determine the

  personality and character traits of others.

  This application is all the more interesting because it can also be used to model an individual profile on a computer network such as the Internet. Indeed, if we couple this application to the different available applications (cosmetics, hairstyles, eyewear, access control, photographic processing, etc.), the system can start from

24 2783949

  each face partly determines the personality of a subject, and thus determines

  a consumer psychological profile.

  The system database is enriched with a set of techniques corresponding to the observation of faces supplemented by those of a person skilled in the art in this field (morphopsychologist, psychologist, expert in purchasing behavior, etc.). Grimming The device according to the invention makes it possible to choose a grimming applied to the face of a child, a teenager, or an adult. In this way, the subject can choose to look like a famous character (comics, cartoons, films, video games etc.) or translate their expression in different fun or festive areas by creating a virtual mask (clown, animals , demons, vampires etc.), see

other choices ...

  Thus the image of the subject's face can be retouched and modified according to the subject. This application can be implemented directly by the process, application and techniques of makeup, ornaments or accessorization of the face

  and capillaries can be directly included in the facial treatment.

  The database corresponding to the aesthetic simulation of the face is completed by a set of rules concerning grimages (famous characters, fun and fun virtual masks etc.), hairstyle, ornamental accessories (false nose, hat etc.) and the color of the hair thus making it possible to simultaneously treat the

  face and hair part of a subject.

Claims (16)

Claims:   1.- Automatic method of producing aesthetically ideal virtual images, usable among other things on the Internet, from the digital capture of the color photograph of a face, characterized in that it comprises the steps of automatically detecting the image of a face; 5. automatically segment the image; automatically determining, during a first acquisition phase, the histograms of the color levels of said segments to determine the fine characteristic features of the face and its skin structure, and the texture of the hair; correct, using an aid and correction system, in an exceptional case where the fine characteristic features of the face and its skin structure and the texture of the hair could not be correctly determined; automatically carry out during a transformation phase, modification or aesthetic improvement of the characteristic features of the image and certain skin colors, of the hair, according to the selection criteria determined by an expert system led by a inference engine by relying on one or more databases enriched with several thousand combinations of shapes, color harmonies and their respective textures; proceed automatically, at the end of the transformation phase, to personal modifications of the new aesthetic image of the face, using a virtual selection tool, making it possible to modify the fine characteristics of this image using of colors and their textures, and of shapes, selected and determined beforehand by the expert system animated by an inference engine based on one or more databases enriched with several thousand combinations of   shapes, color harmonies and their respective textures.   2.- Method according to claim 1, in which after determining the histograms of the color levels of the segments of the face, one proceeds according to the program of the expert system to a recognition of the structures of the face by searching in two preferred directions for places of discontinuity histograms of the color levels which will lead to positioning the different elements of the face in relation to each other, to determine the characteristic features   of the face (nose, eyebrow, eyes, lips etc.) and to generate "geometric" primitives.   3.- The method of claim 2, wherein the contour recognition of the shapes that make up the image is obtained by analyzing the equivalent colors corresponding to each elementary segment of the image, the layout is made using 26 2783949   contour following automata, and in which when the shape or context of the area is delicate, the contour drawing is obtained, by calculating and mathematically applying a parabola on the element previously and approximately recognized in each segment whose will statistically deflect the curve, point by point, according to the average and the coefficient of variation of the color levels of each line of this zone, until each of the points precisely and finely borders the contour, and in which the newly obtained outline drawing is statistically compared again to its virtual "template" model of the object or element   corresponding, until perfect correlation.   4.- Method according to any one of claims 1 to 3, characterized by   the use of expert system decision criteria to choose a technique   cosmetic intended to correct irregularities in skin texture.   5.- Method according to any one of claims 1 to 3, characterized by the   loading into the database of information concerning changes in color of the skin tissue or skin spots corresponding to a set of   dermatological character allowing to suggest the treatment to be carried out thereafter.   6.- Method according to any one of claims 1 to 3, characterized by the   loading into the database associated with the expert system of rules concerning   the corresponding plastic surgery techniques.   7.- Method according to any one of claims 1 to 3, characterized by the   loading into the database associated with the expert system of rules concerning cosmetic techniques, infographics and algorithmic elements relating to calculation of the corresponding textures.   8.- Method according to any one of claims 1 to 3, characterized by the   loading into the database corresponding to the aesthetic simulation of a face of a set of aesthetic rules concerning spectacle frames,   colored glasses and coloring contact lenses.   9.- Method according to any one of claims 1 to 3, characterized by the   loading into the database corresponding to the aesthetic simulation of the face of a collection of aesthetic rules concerning the modeling of hairstyles and the   hair color allowing simultaneous treatment of the face and the hair part.   10.- Method according to any one of claims 1 to 3, characterized by the   loading into the database corresponding to the aesthetic simulation of a face, of a set of aesthetic rules concerning jewelry and accessories ornamental.   11.- Method according to any one of claims 1 to 3, applied as   as a sales support and animation tool allowing to deal at a point of sale or an Internet network, simultaneously simulating the aesthetic image and the proposal 27 2783949   personalized aesthetic advice, characterized by loading into the database corresponding to the simulation and proposal and order management, aesthetic rules corresponding to the field of activity of the manufacturer or distributor, supplemented by a set marketing, promotional, statistics and order and sales management rules.   12.- Method according to any one of claims 1 to 3, characterized by the   loading into the database (s) of information of the characteristic features and dimensions (shapes, spacing of the eyes, height, width, ..., colors etc.) of a   or more known faces for comparison.   13.- Method according to any one of claims 1 to 3, characterized by the   loading into the database of techniques implemented by psychologists, morphopsychologists, proposals for aesthetic advice that have   been determined, as well as observed and recorded reactions.   14.- Method according to any one of claims 1 to 3, characterized by the   loading into the database using make-up and coloring techniques,   of models, and accessories of ornaments.   15.- Method according to any one of claims 1 to 3, characterized in that   that a color photograph of a face supplied on all supports (Argentine paper, slides etc.) and of all formats (identity, portrait etc.) is digitized by a high definition scanner, then is to be retouched and aesthetically improved, even archived , and in which the database is supplemented by cosmetic techniques, infographics and modeling.   16.- Device for implementing the method according to any one of   Claims 1 to 15, characterized in that it includes a multimedia terminal   digital and tele-communicating in which all the databases as well as the archiving system are stored on a remote Internet server which is accessed via the Internet; the device essentially comprising an Intel 486 DX2 or Pentium or equivalent microprocessor, of 60Mhz power and higher coupled to a Ram memory of at least 32 MB and a video memory of at least 4 MB, having a floppy disk reader / writer, a hard disk, a Digital Virtual Disk (DVD), digital storage means, an electronic hard disk memory (hardship disk), of at least 100 megabytes of capacity, at least a digital CCD video sensor of the type used in video communication coupled to its video interface or a color scanner, a reflecting or semi-reflecting mirror with 50% refractive power or not, a color touch screen or not of at least 14 inches in size of diagonal screen, can be plasma technology or active matrix liquid crystal with its specific video / graphic interface allowing a minimum resolution of VGA display of 640 x 480 dots and 256 colors, even 1280 x 1024 dots and a depth 28 2783949   colors of 24 bits (16.7 million colors) or more, a keyboard and its mouse, or a remote control (infrared or radio frequency), or a voice recognition system (voice control), a modem speed of at least 55.6Kbits PSTN or digital ISDN type, a color printer type liquid or solid inkjet, or thermal sublimation, or laser or any other means of color editing, a language called Transmission Control Protocol and Internet Protocol (TCP / IP), a union of processes forming a basis for data transfers over the Internet, navigation software such as Netscape, Internet explorer etc., a set of tools for capturing digital images, an engine for automatic recognition configured to be interfaced and interactively interacted with the expert system of class 0+ or 1 and its database (s), and an editing module which will generate the results obtained, in the form of a color plate of the real picture the initial and the new virtual image and some comments relating to artistic or technical criteria in the   field in which this image could have been created, this multimedia multimedia terminal   communicating being connected transparently to one or more Internet servers which are computers of the Internet or World Wide Web which is a device, on the scale of the Internet allowing, to generate a fabric of very varied publications endowed with a number growing in multimedia and intelligent functions, these servers   with the resources of the TCP / IP language, the description language used for the   presentation of information (HTML) from the World Wide web, HTTP (Hyper Text Transfer Protocol) for data transmission, from one or more expert Class 1 systems equipped with one or more databases, and animated by an inference engine, that is to say a powerful deductive system independent of the database, and coloring and drawing tools for transformation, this expert system being interfaced interactively with the recognition tool, c that is, it is likely to process the information from it and transmit it to   return the orders and the tools he deducted for the transformation of the image.